THE INFLUENCE OF SOCIAL NETWORKING SITES ON HIGH SCHOOL

STUDENTS’ SOCIAL AND ACADEMIC DEVELOPMENT

by

June Ahn

A Dissertation Presented to the

FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EDUCATION)

August 2010

Copyright 2010 June Ahn

 ii

Acknowledgements

This dissertation could not have been possible without the generosity of numerous

individuals. I am forever indebted and grateful for all the support I’ve received

throughout this past year. I particularly want to thank:

My wife Emy, for your unconditional love and support.

My advisor Dr. Dominic J. Brewer, for your steady guidance in all things.

My friend Paul Pasaba, whose software and technical assistance made this dissertation a

reality.

My committee Dr. Guilbert Hentschke, Dr. Janet Fulk, Dr. Richard Clark, and Dr. David

Dwyer for your invaluable mentorship.

Andrew McEachin, Dr. Richard Brown, and Dr. Kathy Stowe for offering a timely

helping hand at various stages of this ambitious project.

And finally I thank all of the superintendants, high school principals, and teachers who

welcomed me into their classrooms and participated in this research project.

 iii

Table of Contents

Acknowledgements ii

List of Tables iv

Abstract v

Chapter 1: Introduction 1

Chapter 2: The Effect of Social Network Sites on Adolescents’ Social 7

and Academic Development – Contemporary Issues and Research
Figure 2.1: Average Number of One’s Active Network 35

Chapter 3: Digital Divide and Social Network Sites – Which Students 60

Participate in Social Media?

Chapter 4: The Effect of Social Network Sites on High School Students 85

– A Cluster-Randomized Trial

Chapter 5: Future Directions for Research on Social Network Sites 121

Bibliography 127

Appendix A 139
 iv

List of Tables

Table 2.1: Digital Divides in Year 2000 22

Table 2.2: Overview of Empirical Studies on SNS 52

Table 3.1: Descriptive Statistics of Key Variables 71

Table 3.2: Cross Tabulations of Youth SNS Usage 74

Table 3.3: Results of Binary Logistic Model on Probability of 76

having a SNS Profile

Table 3.4: Predicted Probability of Having a SNS Profile 79

Table 3.5: Internet Ownership and Internet Access for SNS Users 80

Table 4.1: Power Calculations when Classrooms are the 98

Randomized Unit

Table 4.2: Required Classroom Clusters by ICC x MDE 99

Table 4.3: Characteristics of Classrooms in SNS and Control 101

Groups

Table 4.4: Descriptive Statistics of Dependent Variables and 107

Covariates

Table 4.5: Percentage of Missing Data on Dependent Variables 109

Table 4.6: Results of Random Effects Models for each Dependent 113
Variable
 v

Abstract

This dissertation examines the effects of social network sites on youth social and

academic development. First, I provide a critical analysis of the extant research literature

surrounding social network sites and youth. I merge scholarly thought in the areas of

Internet studies, digital divides, social capital theory, psychological well-being, identity

development, academic engagement, and educational technology to understand how

researchers might examine new social technologies and youth. Second, I examine the

question of digital divide, or whether particular teenage populations do not have access to

online social networks. Using a nationally representative dataset from the Pew Internet &

American Life study, I explore whether there are disparities in teenage access to social

network sites. Third, I report a cluster-randomized trial that was designed to explore

whether social network sites have a beneficial impact when used in high school

classrooms. A total of 50 classrooms, and nearly 1,400 students were randomly assigned

to use an experimental social network site. The results highlight the challenges and

potential of this technology when applied to school contexts.

 1

Chapter 1: Introduction

As the world moves into the second decade of the 21st century, one of the major

markers of this era is the rise and use of online communities. In particular, a paradigm

called Web 2.0 describes recent technologies that focus on networking mass numbers of

individuals into distinct communities over the Internet (O’Reilly, 2007). Social

networking sites (SNS) are online communities designed to connect individuals to wider

networks of relationships, and are one major example of Web 2.0 applications. Sites such

as Facebook have exploded in membership. In a short period of 2007 – 2010, Facebook

estimates that its membership has grown from 50 million to over 400 million users

(Facebook, n.d.). Online social networks are now an integrated part of daily life and

compel questions of how these media platforms affect human development, relationships,

and interaction.

Teenagers are among the most avid users of technology in general and social

network sites in particular (Lenhart, Madden, Macgill, & Smith, 2007b). Recent reports

find that youth spend nearly 10 hours per day using some form of technology, with

socially networked media playing a large role in their daily lives (Rideout, Foehr, &

Roberts, 2010). New technologies are deeply intertwined with adult perceptions about

teenage life. Mimi Ito and colleagues observe that, “Although today’s questions about

‘kids these days’ have a familiar ring to them, the contemporary version is somewhat

unusual in how strongly it equates generational identity with technology identity” (Ito.
 2

Baumer, Bittanti, boyd, & Cody et al., n.d.). The clear finding is that today’s youth are

increasingly connected to the world through socially networked media.

While teenagers are engaged with technology, they are ever more disengaged

from another major component of their lives – school. National analyses find that nearly

30% of high school students do not obtain their diploma on time (Cataldi, Laird,

KewalRamani, 2009). High school completion rates are difficult to measure, but various

independent studies also suggest that nearly one-third of students ultimately drop out of

school (Barton, 2005). When one compares these competing aspects of teenage life –

technology versus education – a simple strategy clearly emerges. Perhaps if educators

begin to integrate social technologies into learning, they will increase student engagement

and achievement in school. Heeding the call of scholars (i.e. Jenkins, 2006; Ito et al. n.d.)

recent policy and research efforts are now racing to develop new social media platforms

and technologies for learning. For example, the Federal Department of Education and

organizations such as the MacArthur Foundation have invested millions of dollars to

build social media platforms, video games, and other digital tools for learning

(Whitehouse, n.d.).

Despite the optimism that social media tools might improve student engagement

and learning, the stark reality is that these new technologies often conflict with the

practices of K-12 schools. Surveys find that the vast majority of school district leaders

believe social technology can improve student learning. However, these same district

administrators typically block student access to online resources like social network sites

(Lemke & Coughlin, 2009). The decision to ban students from accessing social network
 3

sites underscores a major conundrum for educators. Online social networks widen a

students’ access to resources and social support and may have beneficial effects on their

development. Conversely, as student access to the world widens they are inevitably

exposed to potentially negative material and interactions.

The simplest strategy to limit liability and safeguard school districts is to ban

access to these new digital tools. However, such policies neglect the potentially large

benefits of using social media in the classroom. To alleviate this dilemma, educators and

policymakers need a deeper understanding of social media and youth. Several questions

are critical in the area of youth learning with social technologies, including:

• Which youth are using particular social technologies?

• How do they use these technologies to communicate, develop relationships,

socialize, and learn?

• What are the effects of these technologies on youth development?

• What are the effects of these technologies when applied in educational contexts

such as the classroom?

In this dissertation, I explore these questions by examining a particular technology: the

social network site. Communities such as Facebook and MySpace mediate teenage life,

affecting how youth communicate and learn from one another. In addition, social

networks are intertwined into just about every major online community today

(Livingstone, 2008). These factors make SNS a particularly salient focus for evaluation.

Throughout the following chapters I examine different questions surrounding the

phenomena of social network sites and teenage youth. In Chapter 2, I review the extant
 4

research literature that examines SNS. I consider several controversies around SNS and

youth: (a) What kinds of youth are using social networking sites? (b) Does student

participation in these online communities affect their privacy and social relationships? (c)

Do student activities in SNS influence their personal development in terms of self-esteem

and psychological well-being? (d) Does SNS use affect student grades and learning? The

review highlights how research in this field is only just emerging. The few studies that

examine social network sites are mainly exploratory. However, media researchers have a

rich history of scholarship from which to draw new insights. I integrate previous thought

on Digital Divides, Psychological Well-being, Social Capital Theory, and Cognitive and

Social Learning theories to guide SNS researchers in future studies.

In Chapter 3, I present an empirical analysis using a national dataset of teenagers

from the Pew Internet & American Life Project (Lenhart et al., 2007b; Pew Internet &

American Life Project, n.d.). In this study, I ask whether demographic variables such as

education, socioeconomic status, and access to the Internet are significantly related to

whether teenagers participate in social network sites. This line of analysis is typical of

digital divide studies that examine whether particular populations have less access to new

technologies. If new technologies do have positive benefits for individuals, but under-

represented populations do not have access to such tools, there are tremendous issues of

equity and access yet to be addressed (Jenkins, 2006). Most studies of digital divide and

SNS examine adult and college-age populations. I present an analysis of teenage

populations to examine their usage patterns. The results of this paper highlight how the

association between demographic indicators and social media use are weaker in 2007
 5

than seen in earlier studies. Teenage youth of all backgrounds increasingly find ways to

connect with others using social network sites.

In Chapter 4, I consider a question of particular importance to teachers and

education leaders. Through a large-scale experiment, I examine whether using social

network sites in urban classrooms has any causal effect on students’ social capital,

engagement with school, or academic achievement. I build an experimental social

network site that approximates the functionality seen in sites such as Facebook and

MySpace. The key difference in this experimental condition is that the site is private to

two urban, school districts and explicitly for use to exchange educational information.

Working with 50 classrooms and nearly 1,400 students, I utilize a cluster-randomized

trial, where class periods are randomly assigned to use the experimental site. Employing

this randomized trial design, I find that an academic social network site does not

necessarily improve student engagement with their peers, their classes, or increase

student achievement. However, I find exploratory evidence that existing social network

sites such as Facebook and MySpace improve students’ feelings of connection with their

school community. The study offers evidence for one compelling idea: Perhaps schools

should attempt to leverage students existing social networks, rather than block access to

them or impose their own.

In Chapter 5, I outline what is needed in future research about social network

sites, and new technologies in general, to better inform the policies and practices of

schools, educators, parents, and those interested in youth development. In particular,

previous scholarly thought has focused on either a technologically deterministic or social

 6

agency perspective. Technological determinism suggests that a media tool itself affects

social outcomes such as learning, but a long history of research underscores the fallacy of

this philosophy. Scholars who focus instead on social agency, explore how individuals

use new technologies in cultural and social contexts. However, this stream of research

neglects rigorous evaluation of how new media affect youth. Both perspectives in

isolation offer incomplete analyses of how new media, such as SNS, impact youth. I

argue that future researchers must develop and test finer hypotheses that simultaneously

consider the technological affordances of social network sites, the social and cultural

institutions within which SNS are used, and the actual interactions between individuals

that occur in these online communities.

The chapters in this dissertation examine the phenomena of social network sites

and youth through different but complementary lenses: theoretical, descriptive, and

experimental. The summative contribution of these analyses is a deeper picture of how

teenage youth use SNS and its effects on their academic and social development. The

studies show that youth of all backgrounds are increasingly connected via online social

networks. The empirical analyses also show that social network sites are no silver bullet

for improving learning in high school classrooms. The technology itself does not improve

learning, but social media might help students become more connected and engaged with

their school communities. The implications for educators and schools are numerous.

Problems such as student disengagement with education are profoundly significant

issues, and additional research is needed to better understand how online networks

influence youth development and learning.

 7

Chapter 2: The Effect of Social Network Sites on Adolescents’ Social and Academic

Development – Contemporary Issues and Research

The current tools of teenage communication go by a peculiar set of names. Wall

Posts, Status Updates, Activity Feeds, Thumbs Ups, Facebook Quizzes, and Profiles are

some of the ways that youth today communicate with one another. These tools are

features of social network sites (SNS), such as Facebook and Myspace. SNS are part of a

suite of recent web applications, also called social media, which utilize Web 2.0

principles. The term Web 2.0 defines websites that are designed to: (a) rely on the

participation of mass groups of users rather than centrally controlled content providers,

(b) aggregate and remix content from multiple sources, and (c) more intensely network

users and content together (O’Reilly, 2007). People use these web applications to interact

in hyper-aware ways and the scale of this mass communication phenomena is significant.

As of May 2009, Facebook ranked as the 4th most trafficked website in the world and

Myspace ranked 11th highest (Alexa, n.d.). That high school youth are connected to these

global online communities is both a frightening prospect for parents and educators and an

intriguing area for social science research.

Educators and parents in the United States face difficult quandaries concerning

students and SNS. No one denies that youth use these technologies to communicate with

the world, and they do so with high frequency and intensity (Lenhart et al., 2007b). Many

scholars suggest that students learn in new ways using social media and that educators

should embrace these new platforms (Ito et al., n.d.; Jenkins, 2006). In a recent national
 8

survey, the vast majority of school district leaders report that they view social media as a

positive development for education (Lemke & Coughlin, 2009). Nevertheless, 70% of

districts also report that they banned all access to SNS in their schools. Despite the clear

understanding that social media can be vital to student learning and digital literacy,

educators currently struggle with how to comply with regulations like the Children’s

Internet Protection Act (CIPA), as well as overcome general fears about student

interactions in social network sites. To inform both the policy concerns of district leaders

and the local practices of teachers and parents, research is needed to understand how

youth use SNS and what effects it has on their social and academic development.

In this chapter, I consider several key controversies around youth usage of SNS,

and review relevant research that begins to inform these debates. I first define the media

effects framework and outline how this research tradition attempts to understand the

effects of new technologies on social outcomes. Second, I define social network sites and

describe studies that capture how youth use these technologies to develop relationships,

hang out with friends, and learn new skills. Third, the chapter reviews relevant research

that informs several controversies concerning SNS and adolescents. I also connect these

contemporary debates with previous scholarly thought about students’ out-of-school time

(OST) and traditional concerns about the effect of technology on learning. The specific

controversies reviewed are:

• What kinds of youth are using social networking sites?

• Does student participation in these online communities affect their privacy

and social relationships?

 9

• Do student activities in SNS influence their personal development in terms of

self-esteem and psychological well-being?

• Does SNS use affect student grades and learning?

Finally, I outline the overall condition of research on SNS and youth. The current state of

the literature is suggestive of the effects on adolescent social and academic development,

and primarily consists of ethnographic and cross-sectional data. I outline the future

questions that will be critical for the field and suggest relevant methodological directions

to move this emergent research stream forward.

What Can We Learn from a Media Effects Framework?

Many of the controversial questions concerning social network sites ask what

kinds of effects these technologies have on youth development. Given this focus, I work

primarily from a media effects tradition of research. Media effects scholars examine the

outcomes that arise when people use new technologies. Talking about effects engenders

important theoretical discussions that must be laid clear when examining studies. Most

significantly, the term implies a focus on causality. Studies in this framework imply that a

media form, or the features of the technology, causally influences some outcome

(Eveland, 2003). The structure of questions from this perspective is usually in the form

of: Does media affect learning? Does television influence student achievement? Or do

social network sites affect the psychological well-being of adolescents? Media effects

scholars in a variety of fields have quickly come to realize that the answers to these
 10

questions are more complex. Very rarely, if ever, is there a direct causal relationship

between a technology and a social outcome such as learning (Clark, 1983; Clark, 1991;

Schmidt & Vandewater, 2008).

Early media questions often used a technological framework or object-centered

approach (Fulk & DeSanctis, 1999; Nass & Mason, 1990). Such a perspective assumes

and tests whether a technology itself causally affects a social outcome. For example, in

Education a major question of technology research is whether media affects learning.

Education researchers now firmly conclude that media does not affect student learning

(Clark, Yates, Early, & Moulton, In Press). Numerous studies show that the media tool

neither improves nor negatively impacts learning when compared to the same teaching

strategy in the classroom (Bernard, Abrami, Lou, Borokhovski, Wade, Wozney et al.,

2004; Clark, 1983; Clark, 1991). What matters is not the computer, but the learning

behaviors that occur within the software or educational program.

The findings of non-significant media effects on student learning do not mean that

technology has no influence. For example, Richard Mayer (2001) shows through a series

of experiments that the design of a multimedia presentation affects student learning of a

topic. Putting words and pictures closer together on the screen, when they are relevant to

each other, helps students retain more knowledge than when the elements are placed

further apart on the screen. These results do not validate a technological orientation,

where one expects that the computers themselves improve learning. Rather, the

pedagogical strategy of placing relevant words and images together in a presentation

affects cognition. Media researchers understand that the features of a technology afford
 11

certain possibilities for activity. A multimedia video on the computer allows one to

design words and images on the screen, while a computer simulation might guide a

learner using models of real-world cases. A media tool allows for different possible

learning behaviors (Kozma, 1991).

This subtle difference in theoretical orientation is what scholars call an emergent

perspective (Fulk & DeSanctis, 1999) or a variable-based approach (Nass & Mason,

1990). Scholars using an emergent or variable-based approach view technology as a

structuring factor. Features of a technology, not the technology itself, enable and

constrain how one uses that tool. Conversely, social forces such as cultural norms and

behavioral practices influence how one ultimately uses a technology. William Eveland

(2003) offers five characteristics of media effects research that help define how studies

take into account both technological and social variables. Media effects studies have: (1)

A focus on an audience, (2) Some expectation of influence, (3) A belief that the influence

is due to the form or content of the media or technology, (4) An understanding of the

variables that may explain the causality, and (5) The creation of empirically testable

hypotheses.

A focus on audience compels researchers to understand the characteristics of the

youth who use SNS. Knowing who uses, or does not use, social network sites is an

important sociological question for scholars of digital divide. In addition, Hornik (1981)

notes the possible differential effects for disparate populations, “If communication

researchers have learned anything during the previous three decades, it is that

communication effects vary with members of the audience” (p. 197). Current media
 12

studies also focus on the form or content of a technology, and move away from making

black-box comparisons between technologies. Questions that ask whether Facebook is

related to lower grades, or if MySpace is unsafe for children, are broad and uninformative

directions for future media effects studies. Instead, the pivotal questions explore how the

features of SNS enable or constrain behavior. Future media studies about SNS and youth

should not frame questions using a technologically deterministic perspective where one

expects the technology to cause an outcome. Instead, media scholars identify how youth

interaction, communication, and information sharing are the critical variables in

understanding SNS effects on social and academic outcomes. This understanding of

media effects research helps define finer-grained hypotheses of why a tool like SNS

might affect student development, under what uses, for whom, and when.

What are Social Network Sites and How Do Youth Use Them?

When a teenager joins a site like Facebook they first create a personal profile.

These profiles display information such as your name, relationship status, occupation,

photos, videos, religion, ethnicity, and personal interests. What differentiates SNS from

previous media like a personal homepage is the display of one’s friends (boyd & Ellison,

2007). In addition to exhibiting your network of friends, other users can then click on

their profiles and traverse ever widening social networks. These three features – profiles,

friends, traversing friend lists – represent the core, defining characteristics of social

networking sites.

One will notice that SNS also include other media tools such as video and photo

uploading and many websites now employ social networking features. For example,
 13

YouTube is primarily a video sharing service, but users can add others as their friends or

subscribe to a member’s collection of videos. Using boyd & Ellison’s (2007) definition,

YouTube can be included as a type of social network site. As researchers examine the

effects of SNS on social behaviors, they will undoubtedly come across these blurring of

technologies. Sonia Livingstone (2008) notes that SNS invite “convergence among the

hitherto separate activities of email, messaging, website creation, diaries, photo albums

and music or video uploading and downloading” (p. 394). This convergence of

technologies may complicate what one means by the term social network site.

Amidst the sea of what websites can be termed SNS, the technical definition of

social network sites still provides a shared conceptual foundation. Comparing across

common features – i.e. profiles and friend networks – researchers can begin to understand

how various communities co-opt these characteristics to create entirely new cultural and

social uses of the technology. Patricia Lange’s (2007) ethnographic study of YouTube

shows that users deal with issues concerning public and private sharing of video. Some

YouTube users post videos intended for wide audiences, but share very little about their

own identities. Their motivations might be to achieve Internet fame and gather viewers.

Other members upload videos intended for a small network of friends and may restrict

the privacy settings to only allow access to those individuals. The concepts of friend and

social network for these users are entirely distinct.

Dodgeball, an early and now defunct mobile-SNS, is another social network site

that has been studied. In Dodgeball, a user broadcasts their location via cell-phone to

their network of friends:

 14

For example, when users get to a bar or cafe, they can "check in" by sending a
text message to Dodgeball such as "@ Irish Pub." Dodgeball then broadcasts their
location via text message to people in their Dodgeball network. Users can also be
alerted when friends of friends who have checked in to Dodgeball are within a 10-
block radius (Humphries, 2007, para. 5).

The case of Dodgeball highlights how the social network is utilized differently with this

technology. In YouTube, individuals interact with their social network in ways that are

influenced by the video-sharing focus of the site. In Dodgeball, the SNS technology is

used to mobilize and meet up with friends in real-world spaces. The cases of YouTube

and Dodgeball show how friends, friend networks, and profiles are used in very diverse

ways depending on the individual’s motivations and the cultural norms of a particular

online community.

The early research on what is commonly understood as SNS (i.e. Facebook and

MySpace) also explores how teenagers utilize profiles, friends, and friend-networks. The

process of creating profiles has been a major focus of theoretical and empirical

discussion. The common features of profiles include personal information such as one’s

name, location, school affiliation, occupation, and personal interests such as favorite

movies or music. Other vital components of the profile are pictures, videos, and the

comments one’s peers leave on the page. Profiles can be updated at any time and some

sites like MySpace allow individuals control as to how their profile looks. Using

programming techniques, youth frequently apply “skins” to their MySpace profiles that

completely alter the visual design or interface of their pages (boyd, 2008).

Researchers have understood profiles as both an internal and external process of

identity development. Some studies focus on how individuals display and experiment
 15

with their identities through their profiles (boyd, 2008; Liu, 2007; Livingstone, 2008;

Manago, Graham, Greenfield & Salimkhan, 2008; Schmitt, Dayanim, & Matthias, 2008).

Other studies focus on how one’s network of contacts interpret and assess the profile

(boyd, 2008; Walther, Van Der Heide, Kim, Westerman, & Tong, 2008; Walther, Van

Der Heide, Hamel, & Shulman, 2009). The emerging picture is that youth make explicit

decisions to disclose information about themselves on their profiles, and their networks

provide social feedback to those profile displays. Most interactions in social network sites

follow this cycle of self-disclosure and feedback.

The process of displaying elements of one’s identity to the public in SNS is

similar to how humans interact off-line. Donath (2007) observes that, “Whether face-to-

face or online, much of what people want to know is not directly observable” (para. 10).

She contends that much of human interaction consists of signals that communicate the

status and characteristics of an individual. Donath utilizes signaling theory to examine

how one’s self-presentation in SNS develops identity and trust with others. For example,

when a user displays a contact as a “friend” he or she is – in an indirect way – vetting that

that person is in fact who they claim to be. Thus, members who indiscriminately add any

and all friend requests (including fake profiles or people they do not know) in an effort to

seem popular may instead damage their credibility and trustworthiness to others. Among

teenagers, boyd (2008) finds that “it is cool to have Friends on MySpace but if you have

too many Friends, you are seen as a MySpace whore” (p. 129).

In face-to-face settings, identity signals come in forms such as speech, body

movements, facial expressions, and taste statements like fashion. On SNS, these signals
 16

are written textually and visually on the personal profiles of members (boyd, 2008).

Qualitative studies of adolescents and their media use suggest that (1) identity

development is a major element of teenage life, and (2) youth use media such as profiles

and personal pages to express themselves (Livingstone, 2008; Schmitt et al., 2008).

Participation in social network sites also allows youth to participate in different social

groups, and experiment with their identity displays (Manago et al., 2008). In their study

of college students, Manago and her colleagues also note the possible negative aspects of

identity experimentation. They find that the males and females in their sample often

portray themselves in stereotypical, gendered ways. In addition, social comparison plays

a large role in SNS. When profile and friend information is readily available, students

frequently check in on their peers to see how they stack up.

In addition to deciding what to present on a profile, friendship practices play a

particularly large role in adolescent life. Teenagers need to navigate a variety of decisions

including whom to accept as their SNS friend, and setting privacy controls to allow

certain members to view their profile (boyd, n.d.). Studies suggest that the technical

design and cultural norms of a respective SNS influences friendship behaviors. In

YouTube, the motivations for sharing video influence the privacy controls a member sets

on their media (Lange, 2007). Tools such as Dodgeball, that explicitly intends to facilitate

off-line meeting up, have members whose real world networks more closely match their

online contacts (Humphries, 2007). When MySpace introduced its Top 8 function, where

users designated their top friends on their profile, it set off a firestorm of social drama

amongst teens. boyd (2006) noted, “There are tremendous politics behind the Top 8, not
 17

unlike the drama over best and bestest friends in middle school” (para. 32). These

examples highlight how the structure, function, and mission of a respective SNS

influences networking behavior.

Peers exert influence on a youth’s social and academic development. Friends

share information and influence one to behave within group-accepted norms (Ryan,

2000). The characteristics of one’s friends also signal much information about a person.

Donath & boyd (2004) observe some of the ways that individuals’ relationships reflect

their social identity:

In the physical world, people display their connections in many ways. They have
parties in which they introduce friends who they think would like – or impress –
each other. They drop the names of high status acquaintances casually in their
conversation. They decorate their refrigerator with photos. Simply appearing in
public with one’s acquaintances is a display of connection (p. 72).

One question is whether our public, friend networks affect how others’ view us. Just as in

off-line contexts, are we known by the company we keep in our online settings?

Emerging laboratory experiments find that human beings do in fact judge SNS

profiles, and these judgments are quite strong. In an experiment by Walther et al. (2008),

researchers created fake Facebook profiles of a hypothetical college student. The stimuli

profiles differed in several ways: (1) the pictures of the fake person’s friends were either

attractive or unattractive, and (2) the text of the wall posts (comments made by friends)

were either positive or negative towards the profile owner. In total, 8 stimuli were

presented to form a 2 (attractive/unattractive friends) x 2 (positive or negative wall posts)

x 2 (gender of profile owner) design.

 18

Actual college students were recruited and randomly assigned to view one of the

8 stimuli and assess the social and physical attractiveness of the fake, Facebook student.

The researchers found that participants rated the Facebook student as more physically and

socially attractive when his/her friends were more attractive. Positive and negative

comments left by friends also affected how participants rated the Facebook student.

Finally, Walther et al. (2008) found interesting interactions between profile impressions

and gender. Female profiles were rated as more attractive when the wall posts were

positive. However, male profiles were rated more attractive when the wall posts were

negative. In this study, an example negative wall post was “WOW were you ever trashed

last night!” (p. 39). The study suggests that gender stereotypes also emerge quite clearly

in SNS interactions. Females who drink heavily (as implied by friends’ wall posts) were

not seen as attractive, while the same behavior signaled positive attractiveness for males.

In a similar experiment, Walther et al. (2009) manipulated whether identity cues

suggesting extraversion and physical attractiveness were more effective if they came

from the profile owner or from others. Signaling theory and warranting theory suggests

that people would assess other-generated statements as more credible. This hypothesis is

especially likely in social network sites because profile owners can manipulate what

information is presented on their page. Thus, statements from others might be seen as

more credible than statements from the individual. As theory suggested, participants rated

the fake, Facebook students as more extraverted or attractive when others (through wall

posts) suggested as such compared to when the individual (through self-statements on the

profile) asserted this identity.

 19

The two studies offer compelling experimental evidence that what one puts on

their SNS profile is assessed by others and the characteristics of friends are strongly

related to how one is viewed. In addition, the feedback provided by one’s network in a

SNS is influential in the development of social identity. Adolescents use social network

sites in a variety of ways. They disclose personal information about their identities and

tastes on their profiles (Livingstone, 2008). Teenagers must also add or reject friend

requests from their peers, navigating the complicated web of friendship practices (boyd,

n.d.). Finally, the interactions and feedback that one’s network provides in SNS, through

wall posts and comments, show how complex social identity and peer influence processes

occur in these online communities (Walther et al., 2008; Walther et al., 2009).

The majority of current research on social network sites attempts to understand

the phenomena itself. Scholars have been interested in how youth use these technologies,

what cultural practices emerge in these online contexts, and what theoretical implications

SNS have on personal identity and social relationships. The early descriptive and

ethnographic research on youth, Internet, and social media offer rich evidence that (a) the

features of different technologies, for example the MySpace Top 8 case, influence the

social practices of youth within those online communities, (b) SNS are important places

for youth to develop their personal identity, and (c) youth are using technologies like

SNS to mediate their relationships with friends, romantic partners, and broader groups of

peers (Ito et al., n.d.). The questions that parents, educators, and researchers now grapple

with concern the effects SNS have on adolescent outcomes.

 20

Effects of Social Network Sites on Adolescent and Student Development

Discussions about adolescents today differ considerably from the past through the

central role that technology plays in youth lives. Many scholars agree that the:

… values and norms surrounding education, literacy, and public participation are
being challenged by a shifting landscape of media and communications where
youth are central actors. Although today’s questions about “kids these days” have
a familiar ring to them, the contemporary version is somewhat unusual in how
strongly it equates generational identity with technology identity (Ito et al., n.d.,
Introduction).

Similarly, Prensky (2001) coined the popular term “digital natives” to describe the

current generation of youth. He asserts that youth today have grown up entirely

surrounded by technology and their use of it shapes all aspects of their lives.

The technologies that youth utilize today are most definitely new and how

teenagers use them to communicate with each other are clearly novel. Nevertheless, the

technologically mediated activities that youth participate in are similar to past

generations:

Just as they have done in parking lots and shopping malls, teens gather in
networked public spaces for a variety of purposes, including to negotiate identity,
gossip, support one another, jockey for status, collaborate, share information, flirt,
joke, and goof around. In other words, they go there to “hang out” (boyd, n.d.,
para. 2).

Not surprisingly, the apprehensions of parents and educators about SNS are also

comparable to past questions about how youth spend their time.

Social network sites represent a new environment within which to examine

adolescent development and learning. Within this context, I focus on several areas of

concern including: youth characteristics, privacy and safety, psychological well-being,

and learning. Many of these questions have been asked before of other technologies. For
 21

example, Emmers-Sommer & Allen (1999) state that, “The effect of television,

particularly on children’s viewing, was the predominant mass media force in the 1970s”

(p. 479). The authors also note that concern about television and student academic

achievement became a major topic in the 1980’s. In the following sections, I survey the

nascent research about SNS within a media effects framework. I also highlight how

related questions from other media studies (i.e. television) and out-of-school time (OST)

research inform how one can examine SNS effects on students.

What Kinds of Youth are Using Social Network Sites?

Scholars assert that social media represent new skills and ways of participating in

the world. If students are not allowed to use new technologies and contribute to online

communities like SNS, they will not be able to develop the necessary skills and technical

literacy that will be vital in the future (Jenkins, 2006). Stemming from this belief,

researchers continue to wonder whether certain groups of students are systematically

hindered from using new technologies. For example, Seiter (2008) observes that “Young

people famously use digital communications – instant messaging, cell phone texting, and

social networking Web sites – to maintain their social capital, at least with those peers

who can afford to keep up with the costly requirements of these technologies” (p. 39).

The statement succinctly outlines the critical concern of digital divide scholars: (1) there

is an understanding that many people are using technology, (2) the use has some positive

outcome, i.e. developing social capital, and (3) questions remain as to the systemic and

unequal access to the technology.

 22

Concerns about the digital divide arose because early researchers found startling

differences in which individuals used technology. In 1996 scholars generally agreed that

Internet users were “largely politically conservative white men, often single, English-

speaking, residing in North America, and professionals, managers, or students”

(Wellman, Salaff, Dimitrova, Garton, Gulia et al., 1996, p. 215-216). Similarly, digital

divide studies identified gender differences. For example, boys were more likely to own a

computer and use it in their leisure time compared to girls (Volman & van Eck, 2001).

An early U.S. Department of Commerce (2000) publication, Falling through the Net,

highlighted disparities in technology use by race, income, and education level. Table 2.1

reports several of the digital divides found in the report.

Table 2.1: Digital Divides in Year 2000

Demographic Indicator Percent with Access to Computer or Internet

Race
White 55.7%
Asian/Pacific Islander 65.6%
Hispanic 33.7%
Black 32.6%
Income
Over 75k 86.3%
Under 15k 19.2%
Education
BA Degree 64.0%
High School Degree 29.9%
Less H.S. 11.7%

Technology adoption in the year 2000 exhibited stark inequalities. For example, 55.7% of

White households and 65.6% of Asian and Pacific Islander households owned a personal

computer. Only 33.7% of Hispanic and 32.6% of Black households owned a computer.

Such initial reports framed the discussion of digital divides as inequities along major

demographic and social variables: race, income, education, and age.

 23

Nearly a decade later, the picture surrounding access to technology is appreciably

better in the United States. In a recent national survey of families by the Pew Internet &

American Life Project, researchers find that 94% of families (married adults with

children) currently own a computer (Kennedy, Smith, Wells, & Wellman, 2008).

Ownership of computers and Internet access appears widespread, but individuals’ online

behaviors and activities differ considerably. For example, in the United States nearly

every teenager is online, but disparities exist concerning what teenagers do online. Some

write blogs, others create personal web pages, and others create videos (Lenhart, Arafeh,

Smith, & Macgill, 2008; Lenhart et al., 2007b). In terms of participation in social

network sites, Lenhart et al. (2007b) find that 55% of online teens currently have a profile

on a SNS. These current trends in Internet access and participation suggest a critical shift

in how scholars explore digital divides. The term as defined as access to computers and

the Internet may not be as useful as conceptualizing the divide along differences in

participation and skills (Cheong, 2008; Hargittai, 2004).

In a study of the characteristics of college students’ who use of SNS, Eszter

Hargittai (2007) finds few, significant demographic differences between users and non-

users. Gender still appeared as a significant predictor, with females being 1.6 times more

likely to use a SNS than males. In addition, having Internet access through friends or

family also significantly predicted whether a college student used social network sites.

Other traditional indicators such as race and parent’s education had no significant

correlation to the use of SNS. Hargittai’s study underscores the developing trend of mass
 24

adoption of social network sites. Among the college students in her sample, there

appeared to be few systemic inequalities in their access to SNS.

Hargittai (2007) also disaggregated her results based on different social network

sites – Facebook, MySpace, Xanga, and Friendster. She found interesting and significant

correlations between race and particular SNS communities. For example, Hispanic

students were more likely to use MySpace, but less likely to use Facebook compared to

Caucasian students. Asian students were significantly more likely to use Xanga and

Friendster. Such patterns problematize some of the theoretical benefits of social

networks. For example, Wellman et al. (1996) theorized that, “People can greatly extend

the number and diversity of their social contacts when they become members of

computerized conferences or broadcast information to other CSSN [computer supported

social network] members” (p. 225). However, Hargittai notes that if particular groups of

people gravitate to respective communities, offline inequalities may persist online.

Researchers have several opportunities in future digital divide studies. First,

scholars are moving away from questions of access to examine participation and

communication patterns. What matters will not be whether a student has a SNS profile,

but instead what he or she is doing in the social network. Second, one limitation of

Hargittai’s analysis is the lack of a nationally representative sample. Her study examines

a smaller sample of college students. Future studies are needed to examine whether

systematic differences exist in nationally representative data sets of adolescent

populations. Although recent ethnographic studies suggest youth of all backgrounds find

ways to participate in SNS even when they do not own technologies at home (Ito et al.,
 25

n.d.), empirical studies are still imperative. Finally, most digital divide studies offer

cross-sectional, one-time snapshots of user demographics. This data is helpful to

understand the social context of SNS at a particular time. However, user demographics

change rapidly in SNS as millions of new members join every day. Studies are needed to

track the longitudinal trends of these online communities.

Examining the user characteristics of SNS communities is not only helpful to map

trends in inequality, but also to consider finer tuned hypotheses about media effects for

whom. Survey data suggests that female and male youth might use social network sites in

different ways (Lenhart et al., 2007b). Several of the studies reviewed below also suggest

that SNS use has differential effects for individuals with high/low levels of self-esteem or

extraversion (i.e. Steinfield, Ellison, & Lampe, 2008; Zywica & Danowski, 2008).

Research is needed to identify how the cultural, psychological, and cognitive

characteristics of individuals influence their online behaviors, and vice versa. Both

anthropologists and psychologists assert that identity and social development is a prime

component of adolescent life (i.e. Ito et al., n.d.; Livingstone, 2008; Schmitt et al., 2008).

The rapidly changing developmental context of youth offers a promising area for future

empirical work.

Concerns about Privacy and Social Relationships in SNS

A major controversy surrounding social network sites is youth safety and privacy.

Approximately 70% of school districts block access to SNS, and the main reason for this

trend centers on fears about student safety (Lemke & Coughlin, 2009). Legislation such

as CIPA requires school districts to enact policies that safeguard students from
 26

inappropriate content and educate them about safe online behavior (Federal

Communications Commission, n.d.). Given these concerns from educators and parents,

two key questions emerge for SNS researchers. First, researchers have examined whether

students understand the privacy implications of social network sites and also if youth

actively safeguard their privacy online. The second critical question is to examine

whether adolescent relationships with others in SNS offer positive benefits in the form of

social capital.

The early picture concerning youth and online privacy is mainly positive. Nearly

every major social network site offers privacy controls. In fact, “These privacy measures

have given adolescent users a great deal of control over who views their profiles, who

views the content that they upload, and with whom they interact on these online forums”

(Subrahmanyam & Greenfield, 2008, p. 123). Initial research suggests that teenagers

disclose a variety of personal information on their profiles, but they also proactively use

privacy features to manage who can view their content (Hinduja & Patchin, 2008;

Lenhart & Madden, 2007a). Lenhart & Madden report that from a nationally

representative sample of youth, 66% of teenagers limit their profile to particular people in

their network. A cross-sectional study of a college student sample also finds that privacy

concerns did not hinder users’ desire to share personal information on their profiles.

Rather, students used privacy features to control and limit who could view their

information (Tufecki, 2008).

The vast majority of youth, 91% of those who use SNS, report that they utilize the

sites to communicate with already known friends (Lenhart & Madden, 2007a).
 27

Qualitative studies also converge with this finding that U.S. youth mostly use SNS to

interact with friends and not to meet strangers (i.e. boyd, 2008). Descriptive data offer a

different picture of British youth. A survey of approximately 2600 youth in England

found that 54% received friend invitations from strangers either occasionally or

frequently, and that 51% accepted those friend requests (Sharples, Graber, Harrison, &

Logan, 2009). The researchers recommend caution on the part of educators in using

social media, and suggest more education for children on proper online behavior. The

limitation of cross-sectional survey data and descriptive studies is that the results tell us

little about the individual behaviors of youth. The data provide general trends that are

snapshots of one point in time. These trends are prone to change drastically with the

rapidly evolving demographics and use of SNS at any given time.

Additional studies are needed to identify those youth who might be prone to risky

online behavior and why they participate in such activities. While youth generally seem

to understand how to navigate their online privacy, clear variations in individual behavior

exist. Many teenagers are very adept at managing their online information and using safe

practices. A lesser percentage of teenagers may seek out relationships with strangers and

engage in risky behavior. Little research has been conducted on the negative interactions

that might happen in social network sites. However, research on other Internet

applications and youth safety offer lessons for SNS researchers.

In a survey of 412 Dutch teenagers, Peter, Valkenburg, & Schouten (2006)

explored several factors that are related to how much teenagers spoke with strangers

online. They found that younger adolescents were more likely to talk with strangers. In
 28

addition, teens that used the Internet to explicitly meet new friends or to overcome their

own shyness (social compensation) also communicated with strangers more often. The

Taste, Ties, and Time study used a longitudinal dataset of college students in the class of

2009 (Lewis, Kaufman, & Christakis, 2008). The researchers found that female students

tended to have more Facebook friends, and were more likely to have private profiles, than

males. However, the most consistent finding was that students were more likely to have a

private profile if their friends or roommates also used the privacy settings. These findings

validate some important, common sense ideas of youth and online safety. Younger

adolescents or teens that use the Internet for social compensation may be more prone to

strangers online. In addition, those who have fewer peers that understand the privacy

features of SNS may be more likely to have open, public profiles.

The features of the technology tool may also influence the likelihood of

contacting strangers on the Internet. Peter et al. (2006) find that youth who spend more

time in chat rooms talk with more strangers. Ybarra & Mitchell (2008) also find that

“Youth are less likely to be targeted for unwanted sexual solicitation in social networking

sites than they are through IM [instant message] and in chat rooms… and are less likely

to be a target of harassment on social networking sites than they are through IM” (p.

355). A comparison of the technological environments helps explain these trends.

Instant messaging is a tool that is predominantly used by known friends to chat

with each other. Ybarra & Mitchell (2008) note that most cases of bullying and sexual

solicitation happen with others that the teenager already knows. Chat rooms are often

public and un-monitored spaces where multiple people talk synchronously. The
 29

frequency of risky behavior and unwanted interactions is higher in these online forums.

Finally, early research notes variations within different SNS themselves. Dwyer, Hiltz, &

Passerini (2007) find that MySpace users utilize the site to meet new people more often

than Facebook members. Such patterns might be related to the norms of each site.

Facebook originally began as a college-campus based SNS, and thus established

boundaries around one’s social networks (boyd & Ellison, 2007). MySpace began as a

broader and open network. As Facebook has slowly opened its network to high school

students, then to any individual, these dynamics may change. The key point is that

technical and social elements of a respective SNS community may facilitate or inhibit

behavior, and this question requires further examination.

In general, scholars find that youth behavior in social network sites is not as

dangerous as popular fears would suggest (Wolak, Finkelhor, Mitchell, & Ybarra, 2008).

Most youth post much information about themselves in their profiles, but this activity by

itself does not appear to expose adolescents to online predators. Instead, youth who

engage in risky behaviors such as interacting frequently with strangers may more likely

encounter negative outcomes. Teenagers are generally cognizant of the privacy features

of SNS, and utilize them to limit who can interact with them online (Lenhart & Madden,

2007a). While legislation such as CIPA is necessary to guide schools, studies also

suggest that the regulation does little to actually influence students’ knowledge about

Internet safety or behavior (Yan, 2009). Students frequently use SNS outside of school

and learn a great deal about managing their privacy through experience and peers. This
 30

fact does not rule out the need for educators and adults to be involved. The research

reviewed here underscores how vital it is for schools, parents, and teachers to educate

youth about the positive uses of SNS.

Studies about adolescent privacy and safety focus on the potential negative

relationships – with strangers and predators – that can be formed online. However,

scholars also posit that the Internet widens our social networks and provides positive

benefits in the form of social capital (Wellman et al., 1996). Various theorists focus on

disparate elements of social capital theory, which often leads to confusion on the part of

research studies that use the framework (Portes, 1998). Pierre Bourdieu (1986) focuses

his definition on people’s membership to social groups that have cultural and financial

wealth. If one is a member of a group with many resources, he or she can accrue benefits

– financial, cultural, or social – from having that access. James Coleman (1990) defines

social capital in terms of relationship and group norms. Groups that exhibit a high level

of trust have more social capital because they are more likely to help each other. Putnam

(2000) also popularized the term in his book Bowling Alone and SNS researchers have

utilized his ideas of bridging and bonding capital in recent studies. Putnam observed that

diverse social groups provided bridges to new information and ideas, while homogenous

groups most often offered bonding relationships based on social support.

Researchers have linked Putnam’s (2000) concepts to the earlier work of Mark

Granovetter (1983) who used the terms strong and weak ties to describe the quality of

relationships between two people. Granovetter defines weak ties as those who are

acquaintances. These relationships often provide access to information and diverse ideas.
 31

Strong ties are with close friends and family, and like bonding capital, provide social and

emotional support. The diverse perspectives on social capital are worth noting because

SNS scholars often evoke one or more of these definitions; always under the banner of

social capital theory. Portes (1998) offers a more general definition that highlights the

explicit conceptual link between SNS and the theory: “Despite these differences [in

definitions], the consensus is growing in the literature that social capital stands for the

ability of actors to secure benefits by virtue of membership in social networks” (p. 6).

Hypothetically, SNS have the potential to widen a person’s social networks and

provide access to valuable resources, information, and social support (Wellman et al.,

1996). Donath and boyd (2004) observe, “Social networks – our connections with other

people – have many important functions. They are sources of emotional and financial

support, and of information about jobs, other people, and the world at large… Today we

are seeing the advent of social networks formed in cyberspace” (p. 71). Of course, early

Internet theorists (i.e. Wellman et al., 1996) already recognized that social networks were

being formed on the Internet with tools like message boards and forums. Nevertheless,

Donath and boyd highlight the particular potential of a new technology, the social

network site. They also forward a popular hypothesis in Communication research that

online social networks facilitate the creation of weak ties, but not strong ties. The

common theory is that Internet use can widen one’s social networks, but is less effective

in helping individuals develop close relationships.

A series of studies with college students and Facebook test these particular social

capital hypotheses. Ellison, Steinfield, & Lampe (2007) surveyed 286 undergraduate
 32

students to examine whether the use of Facebook was correlated to their levels of

bridging and bonding capital. Using regression analysis, researchers found that higher

Facebook use was positively correlated to bridging social capital. Facebook use was also

positively correlated with bonding capital, combined with other factors such as whether

students lived on campus. This finding highlights the fact that bonding capital, close

relationships, would likely also require face-to-face contact, not just Facebook contact.

Most interestingly, the authors found interactions between Facebook use and measures of

self-esteem (SE) and life satisfaction (LS). For example, among infrequent Facebook

users, high SE students had more bridging capital. Among frequent Facebook users the

comparison reversed. Low SE students had more bridging capital than their high SE

peers, indicating an interaction. College students who have low self-esteem or life

satisfaction appear to benefit more from Facebook usage.

The first study (Ellison et al., 2007) used a cross-sectional dataset. Steinfield,

Ellison, & Lampe (2008) extended the discussion of Facebook and social capital in

another study using a longitudinal data-set. The researchers used a cross-lagged

correlation strategy to simultaneously test competing hypotheses: (a) That Facebook use

at time 1 predicted bridging capital in time 2, and (b) Or the converse hypothesis that

bridging capital in time 1 predicted Facebook use in time 2. They found the correlation

for relationship A to be 0.48 and for relationship B at 0.14, and the differences in the two

correlations were statistically significant. Steinfield et al. conclude that Facebook use was

a better predictor of future bridging capital. The researchers also ran the cross-lagged
 33

analyses on subsamples split by high or low self-esteem. They find that Facebook use had

a higher, positive impact for low self-esteem users.

A study by Beaudoin (2008) used structural equation modeling to examine how

one’s motivation for Internet use was related to the development of interpersonal trust.

Beaudoin, citing Putnam (2000), describes trustful relationships as a component of social

capital. However, taking Coleman’s (1990) perspective on social capital, one could

plausibly define trust itself as positive social capital. Beaudoin finds that Internet users

who are motivated to develop their social relationships online (i.e. a social motivation):

(a) use the Internet more, and (b) have less perceived information overload from the

Internet. These two mechanisms, in turn, significantly predict the level of interpersonal

trust that the participant felt in their relationships.

These three studies suggest opportunities for future research using a social capital

framework. The study by Steinfield et al. (2008) is an important step towards longitudinal

analyses that examine SNS use over time. In addition, they highlight the benefit of testing

competing hypotheses to address the mutual correlation issue of cross-sectional analyses.

Does SNS use develop more social capital? Or do individuals with higher initial social

capital use SNS more often? These are critical questions for non-experimental field

studies. Future correlational, survey research that utilizes longitudinal designs will make

the most impact in the SNS literature.

At a conceptual level, cross-sectional research helps to identify key theoretical

mechanisms. For example, Facebook activity appears to widen one’s social networks

(Ellison et al., 2007). Recent data from Facebook also highlights how SNS increase the
 34

number of people one can interact with regularly (Sandberg, 2009). Figure 2.1 shows the

number of people an average Facebook user interacts with regularly. Reciprocal

communications are conversations between two individuals, and direct communication is

when one sends a message to another user but does not expect or receive a reply. Finally,

Facebook introduced a feature called the “stream” which is a constantly updated listing of

all recent activity by your network. The stream appears on one’s homepage and the

company has found that the feature has increased the number of people an average user

now communicates with (called stream communication in Figure 2.1).

As one can see from Figure 2.1, users approximately double the number of people

they actively communicate with using social network features of the site. For education

scholars the implications are clear. Membership into wider, more diverse, and resource

rich social networks (i.e. social capital) has been associated with a variety of student

outcomes including high school graduation and college going (Dika & Singh, 2002). An

important direction for future research will be to examine whether social capital mediates

the relationship between SNS use and these outcomes.

 35

Figure 2.1: Average Number of One’s Active Network

Source: http://blog.facebook.com/blog.php?post=72975227130

A final insight from cross-sectional analyses is the identification of moderating

conditions. Facebook users who were low in self-esteem and life satisfaction seem to gain

more bridging social capital from using the site (Ellison et al., 2007). In addition,

individuals who are motivated to use the Internet to develop their social contacts, appear

to develop more trusting relationships (Beaudoin, 2008). Perhaps SNS allow students

who may otherwise be less social in school, to find avenues of participation in social life.

Similarly, Education scholars suggest that participation in multiple, diverse social groups

are vital for student success – particularly for traditionally under-represented,

disadvantaged youth (Dika & Singh, 2002; Stanton-Salazar, 1997; Stanton-Salazar &

Spina, 2000). Researchers have a current opportunity to examine how social network

sites facilitate social capital development for youth; particularly its effects on education

outcomes. As of this review, no studies of the educational uses of SNS exist that use the

social capital framework, and the area is ripe for new theoretical and practical impact.
 36

Do Student Activities in Social Network Sites Affect Their Personal Development?

Self-esteem and psychological well-being are the two most common outcomes of

interest in prior Internet and SNS studies. Researchers typically measure self-esteem

using established scales such as the Rosenberg Self-Esteem Scale (used in Ellison et al.,

2007). Psychological well-being often refers to various measures that capture an

individual’s satisfaction with life. Scholars use a variety of scales that include measures

of loneliness, depression and overall life satisfaction (i.e. Kraut, Patterson, Lundmark,

Kiesler, & Mukopadhyay et al., 1998). A key debate among researchers considers

whether higher use of the Internet affects one’s self-esteem and psychological well-being

(Kraut et al., 1998; Valkenburg & Peter, 2009a). Such Internet research informs how

SNS researchers examine psychological well-being.

The often-cited HomeNet study by Kraut et al. (1998) recorded the number of

hours individuals spent on the Internet (using tracking software on the participant’s

computers) and it’s relationship to future measures of social involvement and

psychological well-being. The researchers used path-analysis on their longitudinal dataset

and found that longer use of the Internet was related to increased depression, loneliness,

and smaller social circles. The results suggest that Internet use isolates individuals from

their friends and family, and has a negative impact on one’s psychological well-being.

This effect is known as the reduction hypothesis (Valkenburg & Peter, 2009a).

After the HomeNet project, Internet studies exhibited a wide variety of findings

concerning psychological well-being. Most studies reported no significant relationship

between Internet use and well-being (i.e. Gross, Juvonen, & Gable, 2002). A longitudinal
 37

follow-up to the original HomeNet study also found no long-term effects of Internet use

on loneliness or depression (Kraut, Kiesler, Boneva, Cummings, & Helgeson et al.,

2002). A few studies reported evidence supporting the reduction hypothesis. A study by

van den Eijnden et al. (2008) used longitudinal cross-lagged analysis to examine whether

adolescent use of various technologies was related to future compulsive Internet use,

loneliness, and depression. The researchers found mostly inconclusive relationships, but

did find a significant relationship between instant messenger use and depression. Youth

who used IM more frequently reported higher levels of depression afterwards.

The authors (van den Eijnden et al., 2008) offer various hypotheses for this

finding. Perhaps youth that spend more time on IM developing weak ties, spend less time

face-to-face developing strong ties. Alternatively, socially isolated youth who use IM

may not find the social support they seek in these online environments, thus increasing

their feelings of depression (a poor get poorer hypothesis). These hypotheses remain open

for future research in SNS contexts, and van den Eijnden et al. note that, “Future research

should address the conditions under which adolescents low in social resources either

benefit from or are harmed by online communication with people met online” (p. 664).

The key question for media researchers is to identify those conditions – both

technological and social – that help explain Internet effects.

Valkenburg, Peter, & Schouten (2006) note a major shortcoming of previous

Internet research. Many of the studies treated Internet use as a one-dimensional activity.

In reality, individuals use the Internet for many goals such as information gathering

versus social interaction. In addition, prior studies often do not specify what activities
 38

might affect self-esteem and well-being, and why those specific activities might plausibly

affect these outcomes. Binary specifications of whether a teenager uses a particular

technology or not, will likely prove to be an inconclusive predictor of self-esteem and

well-being. Instead media scholars are now moving towards finer definitions of the

technological environment, activities within that environment, and theoretical

specifications about why those interactions would affect social and psychological

outcomes.

Current media effects studies that examine online interactions instead of broad

Internet use, generally find positive outcomes for youth. Valkenburg et al. (2006)

surveyed over 800 Dutch adolescents to examine relationships between social network

site use, self-esteem, and psychological well-being. The researchers used an overall

satisfaction with life scale to indicate well-being. Using structural equation modeling,

they found an indirect relationship between SNS use to self-esteem and psychological

well-being. Adolescents who frequently used SNS had more friends on the site and also

more reactions on their profile (i.e. friends posted more comments and wall posts). In

addition, the researchers measured the valence of this feedback. Having more positive

reactions on one’s SNS profile was correlated with higher self-esteem, and higher self-

esteem was significantly correlated with satisfaction with life. The results highlight the

emerging sense that use of SNS itself does not cause feelings of well-being. Rather, the

positive or negative reactions that youth experience within the site are the key mechanism

for their social development.

 39

A study by Zywica & Danowski (2008) also illuminates how SNS in particular

might facilitate positive interactions for youth. The researchers surveyed a sample of

college students to explore how their levels of self-esteem and extroversion (sociability)

were related to their ideas of popularity on the site Facebook. They found that high self-

esteem and sociable individuals tended to be more popular both offline and online. Low

self-esteem and sociable individuals had lower levels of popularity. Nevertheless, low SE

users shared more information online, expressed different facets of themselves more

often, and admitted to having done something to appear more popular on Facebook.

The results of this study (Zywica & Danowski, 2008) offer insight into how

behavior, specifically in SNS, might influence self-esteem and well-being. Youth bring

offline experiences – existing friend networks and levels of popularity – to the online

context. Additionally, teenagers who may be lower in self-esteem can use SNS sites to

express themselves and engage in positive interactions. Positive relations accumulate to

portray positive self-images. The study suggests that social network sites give youth

opportunities to find a social niche, express themselves in positive ways, portray their

best selves, and allow different conceptions of popularity. These kinds of interactions

may facilitate positive outcomes in well-being.

Why would earlier Internet studies report negative psychological outcomes, while

recent studies find positive personal development? Valkenburg & Peter (2009a) observe

two changes in Internet behavior that help explain recent, positive results of SNS. First,

the authors contend that when prior studies occurred, “… it was hard to maintain one’s

existing social network on the Internet because the great part of this network was not yet
 40

online” (p. 1). In the late 1990’s, one had less family members and friends online with

which to communicate. Past Internet applications such as chat rooms and forums were

designed to facilitate conversation between strangers. The situation now is starkly

different as teenagers and parents, youth and teachers, all find themselves connected in

social network sites. Adolescents typically do not join Facebook to meet strangers.

Instead, they join because their friends are already members and have invited them to

participate. The Internet is no longer isolating, but connecting people.

The fact that youth frequently encounter known friends and family online

underscores the second change in the Internet (Valkenburg & Peter, 2009a). Web 2.0 or

social media applications are designed to facilitate interaction and communication

through networks. Prior uses of the Internet primarily focused on an individualistic

process of presenting or finding information. Information exchange still plays a

prominent role in online communication. However, current tools make one’s social

network an explicit and visible resource from which to get that information. Social

network sites, through the use of profiles and friend networks, enhance the ways in which

people share information about themselves, their friends, and their lives. Again, the focus

of Web 2.0 applications has been to connect persons rather than information.

Self-disclosure also plays a large role in SNS effects on well-being. Specifically,

researchers posit that when youth disclose and express more information about

themselves the quality of their relationships improves. These positive interactions lead to

improved self-esteem and psychological well-being (Valkenburg & Peter, 2009a;

Valkenburg & Peter, 2009b). This theoretical direction is directly related to scholarly
 41

thought in other frameworks including signaling theory (Donath, 2007) and warranting

theory (Walther et al., 2009). Future studies of social network sites and youth must

consider more detailed measurement of behaviors within the online community. These

interactions – positive, negative, informative, or social – may then better predict

outcomes of youth well-being.

I also note that higher self-disclosure was previously viewed as potentially

negative for youth privacy and safety. Youth, particularly those lower in self-esteem or

sociability, who disclose more about themselves online may develop positive outcomes

of well-being. Conversely, these same youth may be prone to risky behaviors and

negative relationships online. Future studies that consider both outcomes, and further

identify the conditions that interact to lead to positive and negative outcomes, will be

particularly important contributions. The practical impact of such studies is significant.

Parents, educators, and other stakeholders are particularly concerned about if, how, and

why youth may benefit from being online.

Does Social Network Site Use Affect Student Grades and Learning?

Research on social networking sites and learning is particularly slight when

compared to the studies of privacy, safety, and psychological well-being. Of the few

research projects on SNS and learning, some are as of yet unpublished studies (see

Blanding, 2009; Karpinski, 2009a). Most studies of social media and youth education

define learning from a literacy studies perspective (Ito et al., n.d.; Jenkins, 2006). The

literacy perspective focuses on educational practices, such as creating media, rather than

traditional measures of learning such as grades or standardized assessments. This

 42

direction is particularly fruitful to consider how youth’s everyday practices with

technology constitute learning in and of itself, and how these activities are in stark

contrast to practices within school. However, research on traditional measures of learning

– grades and assessments – is even more scant.

To date, two studies exemplify the debate surrounding social network sites, youth,

and learning. A conference paper by Karpinski (2009a) received much media attention

when the researcher found that college Facebook users had lower GPA’s than students

who were not users of the site. The author surveyed approximately 220 undergraduate

and graduate students of one Midwestern University. Using MANOVA analysis,

Karpinski finds that students who were Facebook users typically had a GPA in the range

of 3.0-3.5, while non-users had a GPA in the 3.5-4.0 ranges. The author also finds that

Facebook users were more likely to participate in extra-curricular activities and also

come from science, technology, engineering, and math (STEM) fields. From these

results, the author offers several conjectures and hypotheses. For example, perhaps

Facebook users spend too much time online and less time studying. However, the study

did not rigorously examine counter hypotheses and remains a rather exploratory, basic

attempt to understand the effect of SNS on learning.

Pasek, more, & Hargittai (2009a) quickly published a study to counter the

Karpinski (2009a) paper. The authors note several clear limitations of the Karpinski

study. First, the sample of students is clearly limited. Second, the first Facebook study

utilized few control variables in the analysis. And finally, Pasek et al. took issue with the

liberal conclusions of Karpinski, namely that the original study offered strong evidence
 43

for a negative relationship between Facebook use and grades. Pasek et al. present three

additional analyses that use a larger sample of undergraduate students, a nationally

representative sample of 14-22 year olds, and a longitudinal dataset. The authors utilize

more control variables including race, socioeconomic status, and previous academic

achievement variables. From this analysis, the researchers show that Facebook usage has

no significant relationship to GPA in any of their datasets.

In subsequent responses by Karpinski (2009b) and Pasek et al. (2009b) the

researchers debate the merits of their statistical analyses, including their specifications of

GPA and their participant samples. I refer readers to the particular articles for this

discussion, and instead focus here on more general limitations of this strand of research.

All of the researchers in the debate suggest that the Facebook-GPA relationship is an

interesting avenue for future studies. However, aside from the fact that many youth use

Facebook, there appear to be no substantive theoretical reasons why Facebook use might

influence GPA. Future researchers of social network sites would do well to learn from

previous educational technology and media effects research.

In numerous meta-analyses of media and learning, scholars have firmly concluded

that media itself does not influence learning performance (Clark, 1991; Clark et al., In

press). A particularly salient meta-analysis examined research concerning the effect of

online education on learning (Bernard et al., 2004). The authors firmly conclude that

there is no significant relationship – positive or negative – between online courses and

learning when compared to classroom instruction. Framing the potential learning effects

of SNS as a Facebook-GPA relationship is no more interesting than exploring the

 44

MySpace-GPA, Xanga-GPA, or any new SNS-GPA relationship. In less facetious terms,

using correlational studies to make black-box comparisons between media types will

likely produce an ultimate finding of no significant effects. Some studies may find

positive correlations while others may find negative correlations. This type of discussion

may produce many articles and counter-articles, but the ultimate finding will expectedly

leave media scholars and education practitioners wanting more.

Specifying the SNS to Learning Relationship: Lessons from Out-of-School Time and

Media Learning Research

Fortunately, researchers do not need to re-invent the wheel when developing

frameworks to explore the SNS and learning relationship. Insights from research on OST

learning help to create a theoretical link between SNS and learning. One strategy is to

redefine what one means by the term learning. In their review of OST literacy research,

Hull & Schultz (2001) note that one major contribution of literacy scholars is to

understand the concept of practices. Children’s activities in school – i.e. listening to a

teacher’s lecture, practicing problems on worksheets, taking tests to assess their learning

– can be seen as specialized literacy practices. Formal schooling is designed to teach

students to perform well in those behaviors. However, literacy practices outside of school

may serve very disparate functions than expected in the classroom. In the context of new

technologies, youth today communicate and learn very different practices outside of

school. Creating a YouTube video or engaging in social networking interactions are

different literacy practices than successfully completing a multiple-choice test.

 45

Much of the ethnographic research and conceptual thought on new media and

learning, explicitly or implicitly, take this literacy approach to describe learning (i.e. Ito

et al., n.d.). While these studies are significant as rich descriptions of youth practices,

cultural and literacy perspectives are also vital for media effects scholars. A focus on

practices helps to better specify the theories of why SNS interactions might lead to

traditional learning outcomes. Much of what youth do in SNS deal with personal profile

creation and a virtual hanging out with friends. These behaviors naturally relate to

outcomes such as personal identity, relationship development, and overall well-being.

Subsequently scholars should not be surprised that research on SNS and social outcomes

is more frequent, compared to studies linking social network sites to traditional learning

measures. Researchers interested in traditional academic outcomes such as high school

completion, academic engagement, grades, and test scores must specify what practices

would theoretically improve these outcomes.

The research on SNS, social capital, and psychological well-being offer an

additional link to student learning through the mechanism of academic engagement. The

concept of engagement can be defined in behavioral, emotional, and cognitive terms

(Fredericks, Blumenfeld, & Paris, 2004). Behavioral engagement refers to participation in

academic, social, or extracurricular activities. Emotional engagement describes the

positive and negative feelings students may have towards teachers, peers, and the broader

school community. Cognitive engagement depicts the idea that that a student is willing to

expend the energy to comprehend difficult concepts and learn new skills. As noted in this

review, much of the research on SNS suggests that as students more frequently disclose
 46

information about themselves and interact with their network, they develop higher quality

relationships with others. Education researchers who examine the social context of

learning in areas such as out-of-school time, extracurricular activity, and classroom

climate also find a link between high quality relationships, students’ academic

engagement, and achievement (Eccles & Templeton, 2002; Feldman & Matjasko, 2005;

Martin & Dowson, 2009).

A major hypothesis amongst education scholars is that youth participation in

extracurricular and school activities increases their social connectedness with teachers

and peers (Eccles & Templeton, 2002; Feldman & Matjasko, 2005). This connectedness

is related to increased engagement with school and academics. Numerous studies find a

positive correlation between extracurricular activities with grades and student

achievement on standardized assessments. Engagement has also been related to a lesser

likelihood to drop out of school (Fredericks et al., 2004). These hypotheses are still major

questions for education research and reform efforts. Social network sites offer a new

context within which to observe how relationships influence school engagement, grades,

and student achievement.

Researchers of social network sites have the ability to directly observe how online

relationship networks may facilitate this social learning process. What interactions in

SNS might a researcher expect to affect student engagement? Martin & Dowson (2009)

offer some hypotheses culled from a variety of social learning theories such as

Expectancy Theory, Goal Theory, Self-Determination Theory, and Self-Efficacy.

Expectancy theory and goal theory suggests that one’s peers communicate which
 47

behaviors and goals are of value. For example, a student will value achieving good grades

and set this as a goal, if his or her friends also strive for high achievement. Similarly,

Eccles & Templeton (2002) also suggest that peer groups transmit a social identity that

affects student behaviors. Self-Determination theory proposes that if a student’s

psychological need to belong is met, he or she is much more likely to take academic

risks, explore more ideas, and persist when presented with difficult work. Self-efficacy, a

major part of Bandura’s (2002) social cognitive theory, describes how capable one feels

about accomplishing a task. When teachers, parents, and friends model the kinds of

behavior that lead to academic success (i.e. study habits or information seeking), a

student subsequently feels more capable about achieving success.

Martin & Dowson (2009) observe that high quality relationships with adults,

teachers, and peers impact these social learning mechanisms. These theories also

highlight the educational impact of social network sites. Quality relationships might

allow students to feel more connected to school and thus take academic risks. Other peers

might communicate what goals and behaviors are valued, through their status messages

and wall posts. Finally, students might model positive academic behaviors by posting

their behaviors or sharing information in SNS. These types of interactions begin to

specify how relationship development in SNS may contribute to increased engagement

and learning. Perhaps teachers can utilize social network sites to engage their students,

develop closer relationships, and model positive learning behaviors over time. Such

educational hypotheses have yet to be attempted or tested in formal studies.

 48

Social mechanisms such as psychological well-being and engagement offer a

natural way to link SNS to learning outcomes. In addition, academic content and

information sharing in out-of-school activities predict learning effects. Lauer et al (2006)

conducted a meta-analysis of 35 studies that examined the effect of out-of-school

programs for the reading and math achievement of at-risk student populations. They only

included studies that used experimental or quasi-experimental methods, and find that

OST programs improve students’ reading and math achievement. Supplemental academic

programs appear to improve student achievement in math and reading and social network

sites present an intriguing new setting to deliver new academic content.

Scholarly thought in media learning also suggests that attention and engagement

are important variables for students’ academic learning. Previous media researchers find

that students’ mental effort influences how well they learn from media (Kozma, 1991).

Salomon (1984) found that 6th grade students rated books as a much more difficult

medium to process than television. Students then viewed comparable stories in one of the

two media forms, and those in the book condition performed better in assessments of

their knowledge. Such converging results from disparate research traditions offer a

compelling hypothesis for SNS scholars. The fact that social networking sites are fun and

part of a teenager’s daily life suggests that educators might leverage these tools to

increase engagement with school activities. Interventions that take advantage of this

natural engagement and combine it with challenging academic content may be the

mechanism through which students’ might achieve higher levels of learning through this

medium.
 49

The theoretical discussions from previous OST and media research suggest the

need for finer specifications of behaviors and practices that may lead to learning

outcomes. So what is it about social network sites that might make it a cognitively

beneficial tool? In discussing the benefit of computer supported social networks,

Wellman et al. (1996) state that, “The nature of the medium supports a focus on

information exchanges, as people can easily post a question or comment and receive

information in return” (p. 219). Social network sites potentially allow individuals to

marshal their contacts to solve a problem or find information.

Internet scholars have long understood that “The Net makes possible new types of

decentralised collaborations, enabling large numbers of people to work together on

shared tasks…” (Resnick, 2004, p. 117). The logic is simple. Thousands of individuals

can collaborate to create an online encyclopedia, Wikipedia, because someone in the

world will know and write about a given topic. A user with a large Twitter network can

post a question and expect someone in his or her online circle of contacts to reply with an

answer. Online social networks structure information sharing so that large groups can

solve problems and share knowledge. Resnick suggests that to realize the learning

potential of online networks, researchers must be willing to work with networked styles

of thinking.

While the Internet can leverage mass numbers of users to find a correct answer,

formal education remains an individual process. Our school systems, policies, and

assessments are designed to evaluate whether a particular student possesses some

knowledge. Thus, a critical theoretical question for education researchers is how to

 50

leverage the power of online social networks to benefit the internal learning of the

individual student. Insights from cognitive and educational psychology offer some

recommendations for structuring individual learning from SNS. Kirschner, Sweller, &

Clark (2006) observe that human cognition results from an interplay of long-term and

working memory. Experts in a particular area are proficient because they possess large

amounts of information, mental models in long-term memory, about that situated

knowledge domain. Conversely, one’s working memory is limited in how much new

information a person can process.

The implication of human cognitive architecture on learning is that novice

students need guidance. Kirschner et al. (2006) report that, “Controlled experiments

almost uniformly indicate that when dealing with novel information, learners should be

explicitly shown what to do and how to do it” (p. 79). Online social networks increase the

number of others that one can learn from (Wellman et al., 1996). The larger one’s social

network, the higher probability that someone, somewhere will be proficient in a given

topic. However, the size of a person’s social network is not the only significant variable.

Learners need guidance on what to do and how to do it. This precise definition of the

specific type of information that guides learners helps researchers observe and look for

particular types of learning interactions in SNS. These interactions, not a uni-dimensional

concept of technology use, provide theoretical links to learning.

Finally, SNS researchers can learn much from past studies in television and

adolescent learning. For example, Karpinski (2009a) offers a possible hypothesis that

Facebook users might spend less time studying, thus explaining their lower GPA. This
 51

idea is called the displacement hypothesis, and has been examined by early television

researchers who posited that television took away students’ study time (Hornik, 1981).

Studies of students’ extracurricular activities instead suggest that new media, such as

Facebook, replace or enhance other leisure activities, but do not take away time from

youth (Roberts & Foehr, 2008). The critical question for future studies is not whether

youth use one technology or another, but what kinds of interactions and content they

experience in these virtual settings.

Chapter 2 Conclusions

This paper offers a narrative review of the emerging research surrounding social

network sites and youth. SNS are an intriguing new environment to study because the

technology is such an integral part of teenage life. Given its popularity, parents and

educators have considerable concerns about the effects of SNS on their children and

students. In Table 2.2, I summarize the empirical studies that examine social network

sites. The overview highlights several characteristics of the research base at this current

time. The scholarly literature is clearly just emerging about social network sites.
 52

Table 2.2: Overview of Empirical Studies on SNS

Broader Debate Citation Methodology General Findings

boyd (2006) Qualitative
boyd (2008) Qualitative Social Network Sites
boyd (forthcoming) Qualitative organize information
Hinduja & Patchin Content Analysis through profiles and
(2008) friend networks
Humphries (2007) Qualitative
Ito et al., (forthcoming) Qualitative Youth use them to
Lange (2007) Qualitative express their individual
Lemke & Coughlin Cross Sectional identity, personality,
(2009) and tastes
Lenhart, Madden, Descriptive Survey
What are SNS and How
Macgill, & Smith Teenagers “hang out”
Do Youth use Them?
(2007b) with their friends in
Liu (2007) Content Analysis SNS
Livingstone (2008) Qualitative
Manago, Graham, Qualitative Controlled experiments
Greenfield, & find that users actively
Salimkhan (2008) judge others’ profiles
Walther et al. (2008) Randomized Control and can form clear
Trial impressions of a person
Walther et al. (2009) Randomized Control through their profile
Trial and friend network

What kinds of Youth Hargittai (2007) Cross Sectional Demographic variables

Use SNS? do not predict overall
SNS use

Lenhart & Madden Descriptive Survey Youth mostly use SNS

(2007a) to interact with known
Lewis, Kaufman, & Longitudinal friends
Christakis (2008)
SNS and Teen Sharples, Graber, Descriptive Survey Youth are largely
Safety/Privacy Harrison, & Logan aware of the dangers of
(2009) SNS, but certain youth
Tufecki (2008) Cross Sectional may be prone to risky
Ybarra & Mitchell Descriptive Survey behavior
(2008)
Beaudoin (2008) Cross Sectional Use of SNS appears to
Ellison, Steinfield, & Cross Sectional increase one’s social
Effects of SNS on Social Lampe (2007) capital
Relationships Steinfield, Ellison, & Longitudinal
Lampe (2008)

Valkenburg, Peter, & Cross Sectional Use of SNS is

SNS Effects on Schouten (2006) positively correlated to
Psychological Well- Zywica & Danowski Cross Sectional well-being
Being (2008)

Ito et al. (forthcoming) Qualitative Youth learn different

Karpinski (2009a) Cross Sectional digital skills outside of
Pasek, more, & Cross Sectional & school.
SNS Effects on
Hargittai (2009a) Longitudinal
Learning
Mixed findings on the
correlation between
Facebook and GPA
 53

Most of the research focuses on the phenomenon itself and reveals much about how SNS

is defined and how people use these online communities. Scholars now understand that

social network sites are comprised of particular technical features – profiles, friends, and

friend networks. Youth use these Web applications to express themselves, and “hang out”

with their friends.

I also outline several of the major controversies that surround youth and social

network sites. Parents and educators have tremendous concern about youth privacy,

safety, psychological well-being, social development, and academic performance. While

there is much theoretical discussion about the effects of SNS on youth, the empirical

research that informs these popular debates is currently in a nascent stage. This current

state of the scholarly literature affords researchers a unique opportunity to quickly

contribute new empirical studies about SNS and adolescent development. In addition,

research about social network sites extends previously established research areas. SNS

researchers have currently framed their studies from a variety of perspectives:

Psychological well-being, Social Capital Theory, Signaling Theory, Warranting Theory,

as well as Cultural and Literacy frameworks.

Subrahmanyam & Greenfield (2008) observe that the lines between virtual and

real-world is increasingly blurred for youth today: “… for today’s youth, media

technologies are an important social variable and … physical and virtual worlds are

psychologically connected; consequently, the virtual world serves as a playing ground for

developmental issues from the physical world” (p. 124). The key questions for the field

of youth development and SNS focus on what the emotional, social, and cognitive effects
 54

of using the technology are for adolescents. Empirical studies that examine SNS effects

are few, but fortunately researchers have the opportunity to incorporate insights from a

variety of previous research traditions beyond the theoretical perspectives outlined in the

current literature. For Education researchers, theories of out-of-school time, student

engagement, and cognition inform particular hypotheses about why SNS interactions may

lead to improved academic outcomes.

Methodological Issues in SNS Research

The empirical research on social network sites is largely descriptive and

correlational (see Table 2.2). Future studies in the field will need to develop stronger

evidence of causation, and studying the effects of SNS on youth necessitates multiple

perspectives and methodologies. I take a particular viewpoint in this paper, the Media

Effects tradition, with a stated leaning towards identifying causal mechanisms.

Nevertheless, this review highlights how different methodologies are needed to mutually

inform future studies. Debates about quantitative versus qualitative methodology are well

played out by past scholars (Berliner, 2002; Burkhardt & Schoenfeld, 2003; Feuer,

Towne, & Shavelson, 2002; Slavin, 2002). Instead, the key question for future SNS

researchers is not about methodology, but how each research design contributes to the

questions people have about social network sites.

Qualitative studies about youth and new media will always be vital to the field.

Evaluated as stand-alone empirical studies, ethnographic and other qualitative strategies

offer rich descriptions of culture, practice, and phenomena at a given point in time.

Understood in the context of other research paradigms, the early work of SNS researchers
 55

(i.e. boyd, n.d.; boyd, 2008; Ito et al., n.d., Humphries, 2007; Lange, 2007) sets the stage

for both popular discussion and future empirical work. Berliner (2002) also notes a

common challenge for social science that, “Solid scientific findings in one decade end up

of little use in another decade because of changes in the social environment that

invalidate the research or render it irrelevant” (p. 20). Technology is one area where time

plays a significant role because technologies, and our cultural uses of technology, evolve

rapidly. For this reason alone, descriptive accounts and alternative theoretical frameworks

will always remain vital for media research.

The questions and intent of qualitative approaches are distinctly different than

quantitative and experimental strategies. The critical questions for SNS researchers will

be: How are youth using the technology? In what ways are they interacting with

technology? And what are the critical contexts – cultural, social, and economic – that

shape the use of technology? Future studies in this area will be most helpful when they

particularly highlight trends and changes in youth culture and technology across time,

communities, gender, and a variety of other factors. Studies that further identify the

theories behind youth development, new media, and social outcomes will always be

needed to inform policy, practice, and further research.

The majority of studies in the SNS literature use non-experimental survey

methods. The analysis techniques are varied, but most of the studies utilize cross-

sectional data. A few studies use longitudinal designs (i.e. Lewis et al., 2008; Steinfield et

al., 2008). Strategies such as longitudinal cross-lagged models help to address the issue of

mutual correlation, and push researchers to identify which competing hypothesis is

 56

stronger. For example, do highly sociable youth use SNS more, or does frequent use of

SNS develop higher sociability? Such questions are critical and cannot be answered with

cross-sectional data. Future SNS studies that evaluate trends in use and developmental

outcomes over time offer particularly promising ways to address the mutual causation

issue.

A second challenge of observational data is the presence of selection bias (Slavin

2007). Selection bias occurs when subjects actively choose to participate in a program. In

the case of social network sites, preliminary evidence suggests that youth do decide to

join particular communities (Hargittai, 2007) such as MySpace versus Facebook.

Researchers cannot make strong causal claims about media effects if these selection

patterns are not accounted for in the analysis. For example, Karpinski (2009a) found that

Facebook users in the study sample were more likely to come from STEM fields. If

students in STEM fields typically received lower average GPA’s than their peers in the

humanities, this self-selection bias could explain why Facebook membership was

correlated to lower GPA.

Scholars utilize several strategies with observational data with the aim of reducing

selection bias. Research designs such as regression discontinuity, propensity score

matching, or longitudinal models with fixed effects (Schneider, Carnoy, Kilpatrick,

Schmidt, & Shavelson, 2007) offer ways to control for the influence of individuals and

their self-selection into social networking communities. These strategies for non-

experimental datasets offer several benefits for SNS researchers. It is often difficult to

randomly assign youth in already existing social network communities. In addition, SNS
 57

are built from networked, inter-related individuals. It may be unnatural to randomly

assign one student to a new networked-learning program, but not assign her SNS friend.

The main benefit of approximating random-assignment using observational data is the

ability for the researcher to examine naturally occurring phenomena, but pinpoint causal

relationships using statistical controls.

Quasi-experimental studies match subjects based on similar background

characteristics and may be viable for SNS studies. While the researcher attempts to

approximate randomization on a number of background characteristics, one cannot be

totally sure that the groups are random (Slavin, 2007). Nevertheless, this strategy may

offer scholars the best way to compare socially networked groups of youth. For instance,

a researcher could find two high schools whose students are similar to each other in

socioeconomic status and academic achievement. One school might receive an SNS-

based intervention designed to increase student engagement. This treatment would be

natural because the students are likely SNS friends with others in the school. The other

school, the control group, would also be a natural context because these students will

likely not be SNS friends with those in the treatment school. Thus, the researcher could

make better causal claims about the social network effect between these entire

communities.

Finally, randomized control trials (RCT) offer the strongest strategy to test causal

hypotheses. By randomly assigning subjects to a treatment or condition, each group has a

high probability of being similar on all background characteristics. Thus, any difference

can be attributed to the treatment (Slavin, 2007). RCT for social-network hypotheses are
 58

particularly challenging to consider. The main issue is that social network hypotheses

inherently rely on individuals’ inter-related relationships with their network. The network

is not randomly chosen, individuals actively choose who their SNS friends are. For

example, one might attribute learning gains to the amount of information one’s friends

shared over the SNS. Any random assignment, with a large enough sample, should

equalize pre-existing network characteristics in the treatment and control groups.

However, SNS researchers should (a) pay special attention to these network

characteristics, (b) control for these variables in analyses, and (c) monitor the fidelity of

their treatment.

In a hypothetical example, consider a study where high school students in one

school are randomly assigned to a Facebook-based treatment. The researcher

hypothesizes that more frequent SNS interactions with friends will lead to higher

academic engagement. The random assignments will necessarily cut off many students’

natural social networks. Although random assignment should even out network

characteristics in large samples, the researcher should measure several independent

variables. Individuals in a given group (control or experimental) will have friends who

are also in their group, and friends who were assigned to the opposite condition. If

disparities exist prior to the experiment, ANCOVA or multiple regression analyses that

control for network size will be required. The final measures of engagement in this

experiment may be related to the number of Facebook friends that were also invited to

the treatment. Conversely, scholars should consider that a student in the SNS treatment

might have friends who were not included in the treatment. Perhaps these students would
 59

still share information or affect each other’s academic engagement regardless of the SNS

intervention. Social network hypotheses require that researchers monitor the fidelity of

the intervention and control for the challenges of self-selected relationships.

Future Directions

Researchers of social network sites have a unique opportunity to build a new area

of study, extend previous Internet research, and apply a variety of new theoretical

perspectives that have not yet been explored. Irrespective of the theory SNS scholars

decide to utilize, this chapter also offers four general recommendations for future

research in the media effects framework:

1. Identify the moderating effects of users’ individual characteristics and the

technological affordances of a given social network site.

2. Specify and measure the exact kinds of interactions that occur in the SNS. Move

beyond uni-dimensional specifications of use into more descriptive and

theoretically relevant measures of interaction.

3. Examine how a given construct – i.e. Social Capital or Well-Being – mediates the

relationship between SNS use and students’ education outcomes.

4. Utilize methodologies that make stronger causal claims about the media effect of

SNS.
 60

Chapter 3: Digital Divide and Social Network Sites – Which Students Participate in

Social Media?

Since the release of the first social network site (SNS) in 1997 the growth of these

online communities has skyrocketed. Familiar examples of SNS include Facebook and

MySpace, but there are hundreds of services that cater to a variety of populations (boyd

& Ellison, 2007). These online communities are tremendously popular. For example, as

of December 2009 Facebook reports over 350 million active users (Facebook, n.d.). In

the United States, teenagers who range from ages 12-17 represent a significant and

growing population of SNS users. In a 2007 survey from the Pew Internet & American

Life project, researchers find that approximately 55% of online teens have created a

personal profile compared to 20% of adult Internet users (Lenhart et al., 2007b). Social

network sites now mediate a vast array of communication between students. Research on

SNS offers a ripe arena to explore how youth work, play, and learn in these online

environments. The scholarly literature in this area is nascent, but swiftly accumulating

with descriptive evidence of innovative learning and communication among youth (i.e.

Ito et al., n.d.).

The emerging picture of youth and social network sites suggests that these online

communities mediate a wide variety of peer social practices and learning (Ito et al., 2009;

Jenkins, 2006). Teenage users of social network sites appear to be adept in a variety of

other technologies such as blogging and multimedia production (Lenhart et al., 2007b).

Youth who use SNS appear to be much more comfortable networking with others,
 61

creating their own multimedia, and learning new media literacy skills that are essential in

an increasingly technology-mediated world (see Jenkins, 2006). A great deal of learning

happens between youth in these online communities. Teens are not merely messaging and

flirting in social network sites. They negotiate identity, learn social skills, and become

subject matter experts through peer teaching in topics that pique their interests (boyd,

2006; boyd, 2008; boyd, n.d.; Horst, Herr-Stephenson, & Robinson, 2009). Initial studies

of Facebook usage among college students suggest that individuals also develop more

social capital in these online communities (Ellison et al., 2007). Youth participation in

social network communities may signal the development of important technical skills and

social development.

Not surprisingly, education scholars are now turning their attention to the learning

implications of social network sites (Greenhow & Robelia, 2009; Greenhow, Robelia, &

Hughes, 2009). As researchers begin to explore how youth use social media to learn in a

networked environment, questions about access become ever more critical. The term

digital divide describes the concern about unequal access and participation in new

technologies (Norris, 2001). Youth that are systematically excluded from social network

sites may also lose out on opportunities to develop technical skills, social interactions,

and relationship networks. For example, Seiter (2008) states that teens on SNS can

leverage their social networks for their benefit; well “at least with those peers who can

afford to keep up with the costly requirements of these technologies” (p. 39). Thus,

researchers of media and education must still consider the “fundamental inequalities in

young people’s access to new media technologies and the opportunities for participation
 62

they represent” (Jenkins, 2006, p. 12). Such theoretical discussion gives rise to empirical

questions such as: (1) who is using these technologies? And (2), is there unequal access

to these technologies or new digital divides of participation? Thus, I contribute to the

emerging discussion of social network sites and teenage youth, by revisiting questions of

inequality and digital divide.

In this chapter, I use a nationally representative survey of teenagers in the United

States to examine SNS participation (Pew Internet & American Life Project, n.d.). In the

following sections, I first outline how the term digital divide has been used in the

research literature of the past decade. The discussion highlights how scholars have moved

from questions of mere access to computers to current questions of social participation in

technologies like SNS. I summarize the demographic, socioeconomic, and social

variables that researchers utilize to explore the digital divide. Second, I test whether

traditional predictors of the divide (such as race, education, technology literacy etc.) are

also factors of students’ involvement in social network sites. I utilize logistic regression

to examine whether demographic, socioeconomic, and student-level variables are related

to students’ use of SNS. The results offer surprising insight into the social factors that

relate to youth use of SNS and question some long standing digital divide discussions

surrounding race, education, and access.

Applying Digital Divide Research to Youth and SNS

Scholars loosely use the term digital divide to describe the gaps in access to

technology and technology-related activities among a population. One way to clarify the

term is to clearly define the dimensions along which one describes the divide. For
 63

example, Pippa Norris (2001) uses a three-pronged framework to describe digital divides

as global, social, and democratic:

“The global divide refers to the divergence of Internet access between

industrialized and developing societies. The social divide concerns the gap
between rich and poor in each nation. And finally within the online community,
the democratic divide signifies the difference between those who do, and do not,
use the panoply of digital resources to engage, mobilize, and participate in public
life” (Norris, 2001, p. 4).

As Norris’ definition illuminates, researchers can use the term digital divide to describe

various levels of analysis from global comparisons to smaller communities. In addition,

scholars also consider the unequal access to both the technology itself (i.e. a computer)

and opportunities to participate in media.

Researchers typically examine digital divide based on factors such as gender,

race, or socioeconomic status. For example, an early U.S. Department of Commerce

(2000) publication, Falling through the Net, found disparities in computer access across a

variety of variables. In the year 2000, 55.7% of White households and 65.6% of Asian

and Pacific Islander households owned a personal computer. Only 33.7% of Hispanic and

32.6% of Black households owned a computer. The disparities of computer access based

on income level were even more striking nearly a decade ago. Approximately 86.3% of

households with incomes of seventy-five thousand dollars or more reported having access

to computers while only 19.2% of households making less than fifteen thousand had

access to computers. Finally, Falling through the Net highlighted the inequalities based
 64

on educational level. 64% of households where a member held a bachelor’s degree

subscribed to Internet services, compared to 29.9% of high school graduates and 11.7%

of those who did not complete high school.

Such initial reports such as Falling Through the Net (U.S. Department of

Commerce, 2000) established the understanding that access to technology was highly

unequal. Ethnic minority groups were less likely to use technology. Those from lower

socioeconomic backgrounds also had fewer opportunities to use media tools. Early

studies of computer usage also found that males were more likely than females to use

technology (Volman & van Eck, 2001). As technology becomes a more integral part of

society, the question of whether patterns of inequality continue to exist remains a critical

question.

Current studies suggest that access to technology is becoming quite ubiquitous,

but participation with different media varies considerably. For example, in a recent

national survey of families, researchers find that 94% of families currently own a

computer (Kennedy et al., 2008). While ownership of computers and Internet access

appears widespread, individuals use technology for very diverse purposes and needs

(Kennedy et al., 2008). Teenage populations also exhibit highly differential rates of

media participation. Some youth write blogs, others create personal web pages, and

others create videos (Lenhart et al, 2007a). In terms of participation in social network

sites, Lenhart et al. find that 55% of online teens currently have a profile on a SNS.

These current trends in Internet access and participation suggest a critical shift in

how scholars explore digital divides. The term as defined as access to computers and the
 65

Internet may not be as useful as conceptualizing the divide along differences in

participation and skills (Cheong, 2008; Hargittai, 2004). Recent digital divide scholarship

has explored other factors such as technical knowledge and expertise, continued concerns

about gender, and the role of schools in facilitating access (Aslanidou & Menexis, 2008;

Hohlfeld, Ritzhaupt, Barron, & Kemker, 2008; Livingstone & Helsper, 2007; Tien & Fu,

2008).

What Factors Might Relate to Teenagers’ Use of Social Network Sites?

Previous discussions of the digital divide highlight disparities in technology

access and participation along key demographic variables. Education level, racial and

ethnic grouping, and age were significant lines through which to view the digital divide.

Thus, analyses of new technologies and media usage should control for these variables.

However, computer ownership and Internet access have spread rapidly in the last decade

(Kennedy et al, 2008; Lenhart et al, 2007b). With such wide coverage, there are questions

to whether access divides along demographic variables remain significant indicators.

Emerging evidence from researchers of social network sites suggest that demographic

background is not a significant predictor of access to these online communities.

In her ethnographic research on social network sites, dana boyd (2008) states that,

“Poor urban black teens appear to be just as likely to join the site as white teens from

wealthier backgrounds…” (p. 121). Eszter Hargittai (2007) used logistic regression to

examine a sample of college students and found that age, race, and parent’s education

level did not have a significant relationship to whether they used a SNS. However, both

boyd and Hargittai assert that gender remains a significant variable with which to see
 66

access divides in SNS. Teenagers also exhibit different behaviors online depending on

age. While no studies focus the influence of age on youth social network site

participation, numerous studies highlight how young people behave differently online.

For example, Peter, Valkenburg, & Schouten (2006) find that younger, less experienced

adolescents are more prone to talk to strangers online.

Social network sites, and most forms of social technology, are designed to

promote interaction between individuals. Recent surveys find that girls are significantly

more likely to participate in these social communities (Lenhart et al., 2007b). In addition,

older youth are more likely to participate in social network sites. This pattern could be the

result of many factors. Perhaps younger youth have more restrictions and parental

supervision concerning their online activity. Older youth may have more freedom, and

technical literacy, to explore new media tools. The history of digital divide research, and

most recently studies on participation in SNS, offer the foundation for a set of hypotheses

concerning demographic variables:

H1: Given teenagers’ widespread access to computers and the Internet,

demographic variables such as race and parental education level will no longer

have a significant relationship to usage of SNS.

H2: However, gender and age will still have a significant relationship with social

network site use.

The hypotheses underscore the relative trend of increased computer ownership and

Internet access in the United States. If computer ownership and Internet access is

widespread, demographic variables should not be significantly correlated to online

 67

activities. However, teenage life is characterized by rapid changes in personal

development and social contexts. Factors such as gender and age may still be

significantly correlated to participation in different media such as SNS.

Beyond demographic indicators, subtle social and cultural contexts may play a

larger role in youth media practices. For instance, socialization in families may be

significant factors in children’s access to computers and comfort with using technology.

Recent surveys find that technology is becoming a vital part of family life (Kennedy et al,

2008). Family members who actively utilize technology in daily life may also influence

increased participation in digital media by teens. In addition, type of Internet access may

also be subtle factors for student use of social network sites. In 2008, two-thirds of

households reported having high-speed, broadband Internet access. Thus, a significant

portion of households still uses dial-up connections or no connection. Whether students

access the Internet from home, school, or libraries may affect their participation in online

communities such as SNS. Programs such as E-rate offer discounts to public schools for

network infrastructure and Internet connections (Federal Communications Commission,

n.d.). These programs drastically increased Internet access in schools. In 1994 3% of

classrooms had Internet access, while in 2005 94% of public school classrooms were

connected (National Center for Education Statistics, 2006).

Some students may primarily access the Internet at school rather than home.

Schools often have policy restrictions on Internet usage, block websites, or restrict access

time for non-instructional purposes. Such situations might limit students’ participation in

social network sites. For example, dana boyd (2008) observes that,
 68

Those who only access their accounts in schools use it primarily as an

asynchronous communication tool, while those with continuous nighttime access
at home spend more time surfing the network, modifying their profile, collecting
friends and talking to strangers. When it comes to social network sites, there
appears to be a far greater participatory divide than an access divide (p. 121).

Where students primarily access the Internet could impact their participation in social

network sites. These discussions outline a set of hypotheses concerning types of access:

H3: Parental use of the Internet is positively related to teenagers’ use of social

network sites.

H4: Teenagers with broadband Internet access at home (versus dial-up or no

connections) are more likely to be SNS users.

H5: Teenagers who access the Internet primarily from home (versus any other

location) will be more likely to use social network sites.

These factors then contribute to the evolution of digital divide debates by examining

subtle variations in access and participation.

In this study, I also consider a set of individual level factors. As noted earlier,

teens that are social network users also tend to be super communicators. They are apt to

use a variety of other technologies to communicate frequently with their friends. Thus,

one would expect a significant relationship between a teenager’s intensity of Internet use

and whether they also participate in online social networking.

H6: As teenagers use the Internet and related technologies more frequently, they

are more likely to be users of social network sites.

Initial research on youth and digital media suggest that teenagers use social network sites

mainly to keep in touch with their friends (boyd, 2008; Ito et al., 2009; Lenhart et al.,
 69

2007b). Thus, one might also expect that those teenagers who communicate with their

friends more intensely would also be social network users. Their offline relationships and

activities will continue online as well. These conceptual discussions motivate hypotheses

about the individual factors that teens bring to their SNS membership.

H7: Teens who communicate more frequently with friends and family using

technology, will be more likely to also use social network sites.

These sets of hypotheses are posited by scholars in a variety of fields such as Education,

Communication, and Media Studies. Taken together, they move one’s conceptualization

of the digital divide away from simplistic questions about access to hardware or software.

Instead, they are finer grained reflections on the participation divide. This study

contributes to the evolving discussion of digital divides by (1) considering these new

theories about youth digital divide, (2) applying them to a new online phenomena of

social network sites, and (3) extending initial research in the same vein that has examined

college students (see Hargittai, 2007), to now also consider teenage youth.

Method

Sample and Data

This study utilizes a survey conducted by the Pew Internet & American Life

(PIAL) Project (n.d.). The PIAL conducts regular surveys of various topics focusing on

trends of Internet and technology usage in the country. The data for this study comes

from the Teens and Writing survey that was conducted in 2007. The intent of this

particular survey was to gather evidence about teenagers’ writing habits in relation to

their use of various Internet and social media. From September to November 2007, phone
 70

interviews were conducted from a nationally representative sample of 700 teenagers and

their parents. The survey covered an array of demographic and access questions, but also

gathered detailed data about how teens write in their daily lives and what technologies

they utilize. One subset of questions considered how teens use social network sites, and is

particularly salient for this analysis.

Descriptive statistics of the sample are provided in Table 3.1 below. In all

analyses, this study makes use of the sampling weights provided by the PIAL (see

Methodology in Lenhart, Arafeh, Smith, & Macgill, 2008). The sample weights correct

for oversampling of particular segments of the population and adjust the frequency tables

to better match the population sample of the U.S. Census. After applying the weight, the

total sample for this study is 4,855 (n = 4,855).

 71

Table 3.1: Descriptive Statistics of Key Variables

Mean Frequency Percentage Range

Teenager's Gender - - - 0–1
Female - 2365 48.7% -
Male - 2490 51.3% -
Race/Ethnicity - - - -
White or White-Hispanic - 3223 66.4% -
Black or Black-Hispanic - 581 12.0% 0–1
Hispanic - 784 16.1% 0–1
Asian, Other, and Other-Hispanic - 262 5.4% 0–1
Missing - 5 0.1% -
Parent's Education - - - -
Less than High School - 570 11.8% -
High School - 1650 34.0% 0–1
Some College - 1160 23.9% 0–1
College or Over - 1470 30.3% 0–1
Missing - 5 0.1% -
Parent Uses Internet - - - 0–1
No - 649 13.4% -
Yes - 4206 86.6% -
Teenager's Age 14.52 - - 12 – 17
(1.703)
Home Internet Connection - - - -
Dialup - 1060 21.8% -
Broadband - 3227 66.5% 0–1
No Computer or Internet Connection - 496 10.2% 0–1
Missing - 72 1.5% -
Primary Internet Access - - - -
Home - 3318 68.3% -
School - 631 13.0% 0–1
Other - 274 5.7% 0–1
Missing - 632 13.0% -
Number Technologies Teen Uses 1.68 - - 0–4
(0.99)
Teen Communicates Everyday with Various 1.90 - - 0–6
Technologies (1.59)
Standard deviations for means in parentheses
 72

Analysis

In this study I utilize a binary logistic model to determine the probability that a

teenager has a profile on a social networking site. Social network usage is determined by:

a vector of demographic variables (x’A) that includes Race, Gender, Age, and Parental

Education; Access variables (y’B) that include whether the parent is an Internet user, the

type of Internet access available at home, and where the teenager primarily accesses the

Internet; and finally Communication variables (z’C) that include whether students use a

variety of other technologies (i.e. cell phones, blogs, personal websites etc), and whether

students communicate everyday with a variety of technologies (i.e. cell phones, instant

messaging, telephone etc.). The resulting logistic regression model is:

ln(p) = α + x’A + y’B + z’C + ε

Many of the variables in the dataset are categorical and were dummy coded with

particular reference groups. I note these variables to aid in the subsequent interpretation

and Table 1 illustrates the categorical variables. The reference groups for the following

variables is as follows: Gender reference is female, Parent’s Education is less than high

school, Parental Internet User reference is no, Home modem is dial-up, and Primary

Internet Access is home. The race variable is also noted. The Teens and Writing Survey

structured the race question to capture the diversity inherent in the Hispanic population.

Thus, race is coded into categories such as White or White-Hispanic, Black or Black-

Hispanic, Hispanic, and Other-Hispanic which encompasses the Asian-Pacific Islander,

multi-racial, and other categories. White or White-Hispanic is the reference category.

 73

Findings

In Table 3.2, I present some descriptive cross-tabulations of the digital divide

indicators and the percentage of youth who use SNS. The results offer some intriguing

evidence that the digital divide, at least when it comes to youth participation in social

network sites, is beginning to even out. The term – beginning – is key because while

some demographic indicators exhibit equal access, others still describe gaps. For

example, 60% of both White and Black youth use SNS. However, the rates for Hispanic,

Asian, and other ethnic minorities remain lower in this dataset. Female teenagers are

more likely SNS users than their male peers. Older teens are SNS users at higher rates

than younger teens. Most interesting is the relationship between type of Internet access

and SNS participation. Youth with dialup and broadband access use SNS at equal rates

(approximately 60%), but nearly 68% of youth with no computer or Internet access use

social network sites. These descriptive results fly against what one would expect from a

digital divide perspective.

The descriptive cross-tabulations alone cannot discern the relationship between

digital divide indicators and SNS access. To better examine the relationship between any

one factor and SNS access, one must also control for other important and related

variables. I utilize binary logistic regression to control for the major digital divide

indicators used in this analysis. The results of the regression analyses are outlined in

Table 3.3 below. The variables were entered using a hierarchical strategy. In model 1, I

enter the demographic variables (x’A). The access variables (y’B) are added in model 2.

Finally, I add the communication variables (z’C) to examine the full model. The columns
 74

of Table 3.3 outline the results of each model, and I utilize the full model (model 3) for

interpretation of the data.

Table 3.2: Cross Tabulations of Youth SNS Usage

Percent of Youth with a SNS Profile

White/White-Hispanic 60.0%
Black/Black-Hispanic 60.4%
Hispanic 53.6%
Asian & Other 38.9%

Female 66.5%
Male 49.8%

Ages 12-14 38.3%

Ages 15-17 77.3%

Parent Education
Less Than High School 60.8%
High School 54.2%
Some College 59.9%
College or Above 59.5%

Internet
Dialup 59.2%
Broadband 59.8%
No computer or Internet 67.7%

Primary Internet Access

Home 63.5%
School 51.0%
Someplace Else 60.9%

Hypothesis 1 suggested that traditional digital divide indicators such as race and

parental education would no longer be significant predictors of SNS participation. The

data offer mixed support of this hypothesis. Parental education was largely insignificant

in its relationship to whether a teenager used SNS. Remember that the reference group is

parents with less than a high school degree. Thus, college educated parents did not have a

significantly different influence on their child’s access to and use of social network sites.

However, parents with a high school degree had a positive relationship with the
 75

probability that their child used SNS. Such mixed results lend support to the notion that

social and socioeconomic indicators, such as parent education level, may not be major

predictors of participation divides among youth.

While parent education did not largely define any gaps in SNS use, race was still

a significant predictor for teen participation in social network sites. However, the results

at times go against the common thought around race and digital divides. For example,

Black and Black-Hispanic teens were more likely than White and White-Hispanic teens

(the reference group) to be social network site users. Looking at the odds ratio for this

variable, one sees that the odds that Black/Black-Hispanic teenagers are members of

social network sites are 1.423 times higher than their White/White-Hispanic peers.

boyd’s (2008) assertion that black teens are just as likely as white teens to use social

networking sites holds true in this dataset. While Black/Black-Hispanic teens were more

likely to be SNS users, Hispanic and youth of other ethnic backgrounds were

significantly less likely to be on social network sites. The mixed findings relating

demographic patterns to SNS participation suggest that some digital divides may still

exist, but the relationships are less clear than found in Internet studies of the past.

Differences exist across race and socioeconomic status, but these variations are not

stereotypical.
 76

Table 3.3: Results of Binary Logistic Model on Probability of having a SNS Profile

Model 1 Model 2 Model 3 Odds Ratios for

Model 3
Constant -6.822** -7.330** -7.264** 0.001
(.345) (.385) (.409)
Race- .370** .460** .353** 1.423
Black or Black-Hispanic (.117) (.122) (.130)

Race- -.150 -.242* -.272* 0.762

Hispanic (.109) (.113) (.118)

Race- -.657** -.752** -.880** 0.415

Other (.155) (.160) (.169)

Gender -.855** -.857** -.555** 0.574

(.073) (.075) (.080)

Age .522** .567** .454** 1.575

(.023) (.024) (.025)

Parent Education- .348** .158 .328* 1.388

High School (.129) (.137) (.146)

Parent Education- .304* .043 .011 1.011

Some College (.133) (.144) (.153)

Parent Education .234 -.090 .120 1.127

College or Graduate (.130) (.143) (.153)
Degree
Parent is Internet User - .052 -.024 0.976
(.141) (.152)

Youth has Broadband - .216* .024 1.024

Internet (.091) (.098)

Youth has no Internet - -1.235** -1.199** 0.301

(.191) (.206)

Primary Internet Access - -.404** -.119 0.888

is at School (.112) (.119)

Primary Internet Access - .703** .824** 2.280

is not School or Home (.175) (.189)

Youth uses other - - .481** 1.618

technologies (.044)

Youth uses other - - .359** 1.432

technologies to (.028)
communicate with peers

Standard errors in parentheses

* p < .05, ** p < .01
 77

Hypothesis 2 stated that gender and age would have a significant relationship to

SNS usage amongst youth. The findings clearly support this hypothesis. Male teenagers

were approximately 43% less likely to use social network sites than females. Previous

research often found that male youth were more likely than females to use digital media

(Volman & van Eck, 2001). The relationship with gender appears to shift when it comes

to participation in social network communities. In addition to gender, older teens were

more likely to use SNS than their younger peers. Each year of age increased the

probability of using social network sites by 57.5%. Such patterns might be explained by

parental influences on younger teens. Perhaps older teenagers have less regulations or

rules concerning participation in social network sites.

Parental use of the Internet was not significantly related to youth use of SNS

(H3). In addition, having broadband versus dialup Internet at home did not significantly

influence whether a teenager used SNS (H4). An unsurprising, but still noteworthy

finding is that those teens that do not own a computer or are not connected to the Internet

were nearly 70% less likely to participate in online social networks. Finally, there appears

to be no difference in SNS usage for teens that primarily access the Internet from home or

from school (H5). Surprisingly, teens that report having primary access in other locations

(perhaps libraries, friends’ homes, or most plausibly their mobile phones) were

significantly more likely to be SNS users. Perhaps youth are more likely to use social

network sites away from parental or adult supervision. Another distinct possibility is that

with the rising use of mobile technology such as smart phones and iPods, youth access

their social networks with these devices over their school or home computers.
 78

The results of hypotheses 3-5 suggest that parental influence and Internet access

points are statistically insignificant predictors of the participation divide in social network

sites. It seems that a minimal level of Internet access is sufficient for teens to access

social network sites. The findings support qualitative evidence that youth find a way to

participate in digital media activities, even when they might not own or have access to the

latest technologies (Ito et al., 2009).

Hypotheses 6 and 7 stated that youth who use SNS are also very literate in other

technologies (Lenhart et al., 2007b). The findings support these hypotheses. Teenagers

who used other technologies (i.e. cell phone, computer, blogs etc.) were over 61% more

likely to also use social network sites than those who did not. In addition, youth who

were comfortable using other technologies to communicate everyday with their peers

were 43% more likely to use SNS (see Odds Ratios in Table 3.3). These “super

communicator” teens connect with friends over a variety of mediums, including social

network sites.

Table 3.4 below further illustrates the relationship between technical literacy

practices on the probability of SNS use. The table outlines the predicted probability that a

teen has a SNS profile, given their use of other technologies to communicate everyday.

All other variables are held at their means. The qualifier everyday is important in the

interpretation. A “yes” answer means that a technology is a daily part of one’s life, and

indicates that one has a strong level of comfort with the tool. Table 3.4 shows that as

teens report higher use of different technologies, they are also more likely to be social

network users. Teens who report using one more technology on a daily basis, such as a
 79

cell phone or email, increase their probability that they will also use social network sites

by nearly 3-9%. Youth who are use 6 different technologies are nearly 40% more likely

to also participate in social network sites, compared to those who use no technologies to

communicate with their friends. The analysis underscores the idea that youth, who are

active in online social networks, are also show high levels of literacy with other digital

tools.

Table 3.4: Predicted Probability of Having a SNS Profile

Number of
Technologies used to Predicted Probability
Communicate
"Everyday" Social Network Site User Not a User Discrete Change
0 0.5217 0.4783
1 0.6097 0.3903 0.0880
2 0.6910 0.3090 0.0813
3 0.7620 0.2380 0.0710
4 0.8209 0.1791 0.0589
5 0.8678 0.1322 0.0469
6 0.9038 0.0962 0.0360

Finally, the results presented here highlight the extremely complex social contexts

that youth reside in. For example, the descriptive cross-tabs in table 3.2 show that nearly

68% of youth without a home computer or Internet access SNS, compared to

approximately 60% of youth with home technology access. However, the regression

analyses suggest that having no home Internet access has a significant, negative

relationship to participating in SNS (controlling for other factors). The reality is that

youth perhaps find different ways to find access to social media. Table 3.5 displays a
 80

cross-tabulation of only those youth who use social network sites. They are categorized

by both their Internet ownership and where they primarily access the Internet.

Table 3.5: Internet Ownership and Internet Access for SNS Users

Where Youth Primarily Accesses the Internet

Home School Someplace Else Total
Type of Modem at Home
Dialup 473 112 24 609
(77.7%) (18.4%) (3.9%) (100%)

Broadband 1,606 160 74 1,840

(87.3%) (8.7%) (4.0%) (100%)

None Computer 8 51 70 129

and/or Internet (6.2%) (39.5%) (54.3%) (100%)

The results of Table 3.5 suggest that the overwhelming majority of youth who

have dialup or broadband Internet at home, access the web from their residence. The

majority of youth (93.8%) who either do not own a computer, or do not have Internet at

home, find access at school or in other venues. These youth appear to find their way

online, despite obstacles such as a lack of technology access at home. In addition, using

the Internet in places other than home or school, often the most regulated places, was

positively associated with using social network sites. These combined results offer

several compelling hypotheses for digital divide researchers. First, variables such as

technology ownership may not significantly describe digital divides as in past studies.

Home ownership of computers, in an evolving world where mobile devices and phones

are increasingly connected to the Internet, may not be huge obstacles for youth any

longer. Second, when it comes to social media, youth may participate in these online

communities more often when they access them in less regulated places (i.e. away from
 81

the watchful eye of parents or the schools that ban access to these sites). The patterns of

how youth access social technologies are more complex than seen in previous time

periods, and traditional digital divide indicators are fading in terms of their association

with youth participation in online communities.

Chapter 3 Conclusions

This chapter contributes to past discussion of the digital divide by examining the

current phenomenon of teens and social network sites. I utilize past thought on digital

divides to conceptualize both questions of access and participation. Then I test those

theories to examine the differential participation rates of teenagers in online social

network communities. The results pose new questions to traditional digital divide

conversations. For example, race remained a significant predictor of SNS usage, but in

non-obvious ways. Black and Black-Hispanic students were more likely to participate in

social network sites than their White/White-Hispanic peers. Parental education level,

level of Internet access, and place of Internet access are also insignificant factors in this

analysis. Such findings are in accord with emerging descriptive studies that find that

teenagers are consistently immersed in technology, communicating with their friends

using digital media, and learning from these interactions. Teenagers are getting connected

online and traditional conceptions of digital divide in terms of access may not be

significant in this context. In terms of participating in online social network communities,

youth find a way to get connected.

This study also began to examine social and participation divides. Namely,

teenagers’ overall technical literacy was a significant predictor of SNS membership. Such
 82

findings highlight the importance of allowing youth to explore, use, and gain comfort in

using a variety of technologies. This technical literacy may help these teenagers evolve

and participate in new forms of media. In addition, factors such as gender and age are

significantly related to youth participation in online networking communities. These

findings differ from those of adult populations (Hargittai, 2007), and suggest that youth

are particularly unique sub-populations to examine in future research. Studies that

examine the role of age, gender, and cultural contexts in the ways that teenagers use new

social media platforms promise to be fruitful directions for future research.

The limitations of this study also illuminate future research needs. This analysis

utilizes a nationally representative dataset of youth. However, I only consider a binary

outcome of whether a teenager had a social network profile or not. Recent studies of adult

populations find that divides exist in different kinds of platforms (Hargittai, 2007). For

example, Hargittai finds that white college students are more likely to use Facebook

while Hispanic students were more likely to use MySpace. Future studies that examine

youth populations along different communities may uncover similar divides in SNS

participation. Researchers might also consider more detailed indicators of participation.

For example, teenagers undertake a variety of activities in social network sites. They

write on each other’s walls, send messages, post pictures, comment on each other’s

postings, and interact in numerous ways. Future analyses might consider these degrees of

activity – from minimal to diverse – to uncover a better understanding of the participation

divide.
 83

This chapter also continues the tradition of digital divide scholarship that

examines the relationship between demographic indicators and technology access.

However, as technology ownership and access becomes ever more widespread, scholars

must consider other factors that contribute to why a particular population will utilize a

new media tool. Social and cultural indicators may better predict why youth use

particular online communities, beyond indicators such as race or socioeconomic status.

Already, emerging studies suggest that there is a relationship between SNS use and

factors such as self-esteem, popularity, and trust (Beaudoin, 2008; Ellison et al., 2007;

Zywica & Danowski, 2008). These psychological and cultural factors may ultimately

provide more informative when examining why individuals use or do not use a particular

technology.

This chapter illuminates the need for more detailed data concerning teenagers and

their use of social network sites. Scholars must keep one eye on continual questions about

digital divides. As new media emerge, questions of who is accessing and using new

technologies will remain foundational concerns. Digital divide research considers issues

of equality and opportunity for using new media. However, understanding the user

characteristics of new technologies is also imperative for other research endeavors.

Studies that take into account selection effects and patterns of participation, promise to

offer finer insights into the social and educational effects of social media on youth.

Finally, continued research is needed because trends in media use change quite rapidly.

This study only considers a cross-sectional dataset, which was collected in November of

2007. Patterns of media usage undoubtedly changes quite rapidly each year. For example,
 84

Facebook saw an increase in membership from 100 million members in 2008 to over 350

million in 2009. Such widespread adoption of social network technologies suggest that

traditional digital divide indicators will be less meaningful in the near future. Instead,

finer understanding of the social and cultural trends amongst youth and technology are

needed.
 85

Chapter 4: Effect of Social Network Sites on High School Students – A Cluster-

Randomized Trial

In a world in which knowledge production is collective and

communication occurs across an array of different media, the capacity to
network emerges as a core social skill and cultural competency (Jenkins,
2006, p. 49).

At the present time, social networking software is under fire from adult
authorities, and federal law makes it more difficult to access and deploy
these tools in the classroom (Jenkins, 2006, p. 51).

The two quotes from media scholar Henry Jenkins succinctly convey a set of

conflicting developments in K-12 education today. Current information technologies,

what O’Reilly (2007) terms Web 2.0, use the network as its main organizing paradigm.

Social network sites (SNS) such as Facebook and MySpace allow individuals to post

personal profiles and connect with their friends. Twitter encourages users to post short

updates of their lives for their network to see. These examples of social media have

permeated every aspect of society, revolutionized the way companies do business, and

altered how individuals communicate with their friends and colleagues (O’Reilly, 2007).

Scholars also assert that social technologies offer tremendous potential to change how we

educate students in this media landscape (Jenkins, 2006; Ito et al., 2009). No longer do

youth only learn by sitting in classrooms and listening to teacher lectures. They learn

from their peers and the larger community as they text, twitter, network, search for and

create their own knowledge through the Internet. Media scholars observe that youth learn
 86

in these contexts and are also deeply engaged with learning with these technologies

(Greenhow, Robelia, & Hughes, 2009).

Not surprisingly schools are under increasing pressure to utilize new technologies

to engage students and increase achievement. Recent polls of school district leaders find

that nearly 75% view social technology as having a positive impact on students’

communication skills, relationships between students and teachers, and quality of

schoolwork (Lemke & Coughlin, 2009). These K-12 leaders also hope that social media

can help keep students engaged with school. Despite such optimism, 70% of school

districts enact policies that ban student access to social network sites. Districts would

rather manage and control technologies in-house, than allow students to use the sites they

frequent such as Facebook and MySpace. The issues surrounding web 2.0 tools in

schools center on questions of student safety and academic use. District administrators

are deeply concerned about poor student behavior such as bullying, access to sexually

explicit material, or social media being a distraction from schoolwork. Therein lies the

conundrum for educators today. Social technologies capture the attention of students, but

questions remain concerning whether these new tools have a beneficial impact on student

outcomes.

In this paper, I address these concerns by examining a popular technology – the

social network site. Examples of these online communities include Facebook and

MySpace, which are among the most visited sites in the world (Alexa, n.d.). First, I

outline how SNS function and summarize several common theories-of-action that may

link to student outcomes. For example, SNS are commonly assumed to help individuals
 87

develop relationships with peers and feel connected to their communities (Beaudoin,

2008; Ellison et al., 2007). Would using such technologies in a school context improve

student relationships with others and motivation for school? Second, I report an

experiment that was conducted in two urban school districts. A social network site was

built for private use in high school classrooms. Using a cluster-randomized design, a total

of 50 classrooms were randomly assigned to use the SNS for a 6-week period. Students

were assessed on measures of social capital, school engagement, and course performance.

Third, I report the results of the experiment. The school-based social network site was

poorly received and had little impact on student outcomes. However, students’ use of

existing social networks on Facebook and MySpace had positive and significant

relationships with social capital. Finally, the paper offers recommendations and future

directions for research that examines the effect of social technologies in education

contexts.

Theoretical Framework

Technology is an integral part of teenage life. A recent report from the Kaiser

Family Foundation (Rideout, Foehr, & Roberts, 2010) finds that teenagers spend over 10

hours per day using some form of media – including television, music, mobile devices,

and computers. Youth also increasingly access their media through networked, online

platforms such as social network sites. As noted earlier, boyd & Ellison (2007) define

SNS as having three underlying features. First, users create profiles that describe

themselves to others. Second, profiles display the friends of that individual. Finally, users

can click on profiles and friends to traverse ever-wider social networks. This simple set
 88

of features facilitates frequent communication between large numbers of individuals.

Most research on SNS show that social interaction in these online communities works

much the same as offline relationships. Individuals learn about each other through their

profiles, develop judgments on others’ personalities and social traits, and interact with

their networks (Walther, Van Der Heide, Kim, Westerman & Tong, 2008; Walther, Van

Der Heide, Hamel & Shulman, 2009). The online medium intensifies and broadens the

scope of these interactions.

One can find elements of social network sites in just about every major Internet-

based community or technology. For example, YouTube is a SNS that distributes online

video. Facebook is a SNS where users keep in touch with their friends. MySpace has a

niche as a community of musicians and their fans. All of these online communities

revolve around a social network where users connect with one another and share media

from text messages to photos and videos (Livingstone, 2008). From a simple foundation

of profiles, friends, and networks are Internet technologies that encourage individuals to

interact, share, and communicate in a rapid and constantly connected way.

Media scholars theorize that this networked sharing and interaction represents

new learning opportunities for youth (Ito et al., 2009; Jenkins, 2006). For example, youth

write their own blogs, create their own videos and media, and share what they know with

their peers. From this scholarly base, recent initiatives such as the federal government’s

Educate to Innovate fund and the MacArthur Foundation’s Digital Media and Learning

grants have invested millions of dollars to develop social technologies for learning

(Whitehouse, n.d.). The hope is that by engaging students with social technology, and by
 89

encouraging a learning process of information sharing and networking, we might improve

student engagement and learning.

Despite such optimism for social media and learning, little empirical evidence

exists around the effects of these tools on student outcomes. Research in

communications, sociology, and developmental psychology suggest that social network

sites help teenage youth create social capital and develop their personal identities (i.e.

Ellison et al., 2007; Livingstone, 2008; Manago, Graham, Greenfield, & Salimkhan,

2008; Schmitt, Dayanim, & Matthias, 2008). Education researchers are only now

beginning to turn their attention to the role of social technologies in education

(Greenhow, Robelia, & Hughes, 2009). Most research of SNS to date often (a) consider

college student populations, with few studies that examine K-12 youth and (b) offer

limited theoretical understanding for how social technologies affect education-related

outcomes. For K-12 policymakers, teachers, and school leaders interested in how SNS

affect their students, there is a slight research base to inform education decisions. For

education researchers, one must look to studies in other disciplinary and theoretical

foundations to extrapolate the potential of social network sites on educational outcomes.

The Case for SNS Effects on Students’ Social Capital

The clearest potential benefit of social network sites is the ability to develop

relationships with friends, family, and colleagues. Similarly, scholars have posited that

participation in SNS might affect a person’s social capital. Different theorists focus on

particular aspects of social capital theory (see Bourdieu, 1986; Coleman, 1990; Lin,

1999; Putnam, 2000). However, Portes (1998) notes that, “Despite these differences [in
 90

definitions], the consensus is growing in the literature that social capital stands for the

ability of actors to secure benefits by virtue of membership in social networks” (p. 6).

Social network sites are a major place for teenagers to hang out, communicate with

others, and develop relationships (Ito et al., 2009). The intuitive theory is that as youth

develop better relationships, social capital, they may also accrue other benefits associated

with these networks.

The benefits of having positive relationships with others are numerous. In the

education literature, social capital is positively related to important education outcomes.

For example, students with higher levels of social capital are less likely to drop out of

school, achieve higher scores on standardized assessments, and expend more effort on

homework (Dika & Singh, 2002). Early studies of social network site use among college

students find a positive correlation with social capital development (Ellison, Steinfield, &

Lampe, 2007; Steinfield, Ellison, & Lampe, 2008). Students who use Facebook

frequently develop more acquaintances over time that they might turn to for information

or advice. Researchers also find suggestive evidence that SNS benefit students with lower

self-esteem and life satisfaction at higher rates than those already high in these traits.

Finally, recent studies of Internet usage suggest that individuals who go online to interact

with others tend to develop more trusting relationships (Beaudoin, 2008).

From a social capital framework, using SNS in school contexts may offer positive

benefits to students. It is plausible that utilizing social network sites in schools may

encourage students to build relationships with peers, teachers, educators, and other

school-based stakeholders. Another potential effect of SNS is that individuals develop

 91

more relationships with others online, versus in their offline contexts (Williams, 2006).

Perhaps networking sites have less effects on relationships one already has, but aids in

widening one’s social network in online communities. The first two hypotheses in this

study consider the effect of SNS on youth social capital in both their offline, school

community and in their online relationships. The first hypothesis asks whether using SNS

leads students to develop stronger relationships with others in their school community.

H1: Students who use social network sites will have higher social capital in their

school relationships than those who do not.

Similarly, students who use SNS may also develop relationships with others online. Thus,

the second hypothesis states:

H2: Students who use social network sites will have higher levels of social capital

in their online relationships than those who do not.

The Case for SNS Effects on Students’ School Engagement

Keeping students engaged with school is cited as the most prevalent hope for

social technologies (Lemke & Coughlin, 2009). However, no studies to date empirically

evaluate whether using technologies like SNS positively impacts K-12 student

engagement. Emerging ethnographic studies suggest that students are quite engaged with

technology (Ito et al., 2009). In addition, youth are very engaged with the topics and

skills they encounter while online. The natural extension of these findings is to suggest

that using social media in schools will increase student engagement with academic

classes. However, learning a hobby on one’s own time creates a very different

motivational and contextual situation than learning in a formal classroom. Questions

 92

remain as to how a technology like SNS will impact students’ engagement when applied

in a classroom context.

Academics are only one of the ways that students feel engaged with their school

communities. Fredericks, Blumenfeld, & Paris (2004) identify three general types of

engagement: cognitive, behavioral, and emotional. Students may be interested in the

subjects they learn, feel connected with friends at school, with teachers with whom they

have supportive relationships, or may feel satisfied or happy with their school experience

for a variety of factors. In these psychosocial areas, previous research on the Internet and

social technology offers circumstantial evidence that these tools might improve student

perceptions of their school. As noted earlier, social network sites might help students

create positive relationships or social capital with peers and teachers. Such positive

connections might also be related to being more engaged with the school community.

Internet researchers also find that interacting with others online is related to

higher levels of self-esteem and psychological well-being. The literature on

psychological well-being often refers to levels of life satisfaction or happiness.

Valkenburg, Peter, & Schouten (2006) surveyed over 800 Dutch adolescents and find that

those who used SNS had higher numbers of friends online. These teenagers also had

higher frequency of interaction with friends in the social network, which was related to

higher levels of self-esteem and life satisfaction. Furthermore, in a study of college

students, researchers found that students who were low in self-esteem used SNS to

express themselves and engage in positive relationships with their peers (Zywica &

Danowski, 2008). These studies suggest that social network sites allow students to find a
 93

social niche, express themselves, and develop positive relationships with their peers.

These behaviors lead to higher levels of satisfaction or self-esteem.

SNS research on self-esteem and psychological well-being proposes that these

online communities help youth develop positive relationships with others. The concepts

of self-esteem and well-being provide a link to the idea of emotional engagement

proposed by Fredericks et al. (2004). If students use SNS to develop better relationships

with their peers and teachers, perhaps they will also exhibit positive perceptions of their

school community as well. The third hypothesis in this study is:

H3: Students who use SNS will exhibit higher levels of emotional engagement

with their school community than those who do not.

The Case for SNS Effects on Students’ Academic Achievement

The ultimate question for educators and school district leaders is whether using

new technologies like SNS improves student learning. The few studies in this domain

offer scant evidence or theoretical understanding for how social network technologies

might impact student achievement. The first study to consider the relationship between

SNS and student achievement appeared in a conference paper by Karpinksi (2009a). The

researcher surveyed approximately 220 undergraduate and graduate students and found

that college Facebook users had a lower GPA than non-users. She concludes that perhaps

students who use SNS spend more time socializing, less time studying, and thus exhibit

lower academic performance. Pasek, more, & Hargittai (2009a) published a response

using a nationally representative dataset of 14-22 year olds. The authors utilized more

control variables that are correlated to achievement including race, socioeconomic status,
 94

and previous academic achievement. The authors find that Facebook usage does not have

a significant relationship to achievement.

All of the authors in this particular debate acknowledge that their correlational

analyses are limited in several ways (Karpinksi, 2009b; Pasek, more, & Hargittai, 2009b).

The studies do not account for self-selection into particular online communities. For

example, studies of who uses different social network sites find distinct demographic

differences in users. Hargittai (2007) finds that in a college population, Caucasian and

higher SES students were more likely to use Facebook. Minority and lower SES students

were more likely to use MySpace. Youth will also likely separate into online

communities along the lines of race, SES, or previous academic achievement. For

example, Karpinksi (2009a) finds that college students who use Facebook had lower

average GPA. However, the Facebook users in this sample were highly correlated to

being in science and engineering majors where GPAs are typically lower than those in the

humanities.

The second issue in these early achievement studies is the misspecification of

SNS usage. Individuals use SNS and the Internet for varying reasons (Valkenburg &

Peter, 2009), but interactions are mostly social rather than academically related. It is not

entirely clear why using Facebook or any SNS would relate to academic achievement.

The previous discussion of social capital and engagement offer tangential influences on

achievement. Perhaps students who are better connected to their school communities will

also exhibit higher achievement measures such as GPA. However, the early SNS studies

on student learning do not control or measure for specific types of usage. One possibility
 95

is that if students network for academic related goals, one might observe increases in

achievement levels.

Social networks impact relationships and social environments, but also are a

source for information exchange. As the Internet began to be used by the mainstream

population, Wellman et al. (1996) observed that, “The nature of the medium supports a

focus on information exchanges, as people can easily post a question or comment and

receive information in return” (p. 219). Similarly, social network sites may provide a

platform for students to share information and assist each other during the learning

process. If a student does not understand a topic or a homework question, they can post it

to their online networks with the hope that their peers may help them. Such academic

uses of SNS may provide the link to student achievement. To consider this possibility, the

fourth hypothesis suggests that:

H4: Students who use an academic focused social network site will exhibit higher

course grades than those who do not.

School Based versus Organic Social Network Sites

While there are many potential benefits of SNS for adolescent youth, the majority

of school districts ban access to these technologies. Furthermore, district leaders would

rather build and manage their own tools than allow students to use popular versions such

as Facebook and MySpace (Lemke & Coughlin, 2009). There remain some questions as

to whether this strategy, school-based versus already existing networks, will have any

effects on students. Media scholar danah boyd (2009) theorizes that introducing tools like

Facebook and MySpace to classrooms might be detrimental for students. She argues that
 96

these tools serve social purposes outside of school that would dramatically alter when

used within school. Examples of these conflicts emerge consistently in media reports of

SNS use. For example, a Florida student was suspended from school in 2007 for creating

a Facebook page where she and other students criticized a particular teacher (Phillips,

2010). These conflicts occur because schools, students, teachers, and parents do not yet

understand how to meld the potential of social media in school.

The response from education administrators has been to block access to these sites

and attempt to create their own versions for academic use. Little is known about what the

effect of this strategy may be. For example, an academically focused SNS might

encourage students to share information and help each other with class work. There is

little evidence as to whether students already do this on their existing networks in

Facebook or MySpace. Perhaps a school-based network site will have a more direct and

positive impact on student achievement than more socially oriented sites. Conversely,

students might view the school-based network as uninteresting or less engaging,

preferring instead to utilize the networks they have already created organically. In this

case, schools might be better served to leverage students’ already existing networks to

encourage academic engagement and achievement. Opposite of boyd’s (2009) assertion,

it might be better for schools to figure out how to safely utilize Facebook or MySpace to

reach students in their already present communities. Beyond personal declarations and

anecdotal accounts, no studies examine these possibilities. Thus, this study also considers

an exploratory question:
 97

R1: Would a school-based social network site offer positive benefits to students

compared to use of already existing social networks in Facebook and MySpace?

Methodology

I conducted a cluster-randomized controlled trial (CRCT) in 2 urban school

districts. A private social network site was built for academic use in the 2 districts.

Classrooms were then randomly assigned to treatment (using the SNS) or control

conditions. A cluster-randomized design within an actual K-12 field setting was chosen

for several reasons. The CRCT design balances the need for experimental evidence and

maintains a level of external validity. Randomly assigned subjects ensure that treatment

and control groups are equal in important background characteristics. Thus, the effects of

an intervention are better attributed to the experiment over extraneous factors.

While randomized control increases the internal validity of a study, there are

particular challenges to utilizing the design in social network sites or in education

contexts. It is difficult to randomly assign youth to a social networking condition such as

Facebook or MySpace because many students are already members of these

communities. However, the school districts in this study ban access to these SNS and

were interested in implementing a within-district solution. This context created a natural

situation where a new tool could be implemented and evaluated using a randomized

control trial. In addition, many education interventions cannot randomly assign individual

students because they are already clustered into classrooms and schools (Hedges &

Hedburg, 2007). This situation requires the researcher to consider clusters, in this case

individual classrooms, as the unit of randomization.

 98

Power Analysis

Cluster randomized trials require a significant number of clusters to ensure

enough statistical power to detect effects. The number of clusters is the main contributor

to the effective sample size and statistical power results (Donner & Klar, 2000).

Similarly, attrition of clusters dramatically impacts the ability to detect effects in CRCT

designs. I conducted initial power analyses to determine required sample and minimal

detectable effect (MDE) sizes using the multi-level framework presented by Schochet

(2005; 2008). The MDE units are in terms of standard deviations. For the estimates I

utilize the following assumptions: an intra-class correlation (ICC) of 0.15, alpha of 0.05,

an average class size of 27, and power at 80%. The required sample sizes and MDE are

presented in Table 4.1 below. The use of regression models with baseline covariates also

improves the statistical power of cluster-randomized designs (Donner & Klar, 2000;

Schochet, 2008). Thus, I also include MDE and sample size calculations assuming the

use of baseline covariates, under the assumption that the R2 of the regression is 0.20 or

0.50 (Schochet, 2008).

Table 4.1: Power Calculations when Classrooms are the Randomized Unit

Required Number of Minimal Detectable Effect Size With Baseline Covariates With Baseline
Classrooms (n) (MDE) (R2 = 0.20) Covariates
(R2 = 0.50)

35 0.41 0.37 0.29

50 0.35 0.31 0.25
100 0.25 0.22 0.17
270 0.15 0.13 0.11
600 0.10 0.09 0.07
 99

Many social science domains consider an effect size of 0.10 - 0.40 standard

deviations to be a minor effect. However, major educational interventions such as class

size reduction report effect sizes as small as 0.15 (Finn & Achilles, 1999). The definition

of “small” in conventional social science contexts can be significant in education settings.

Given these parameters, Table 4.1 underscores the challenge of conducting cluster-

randomized trials. To achieve an effect size of 0.15, one would require 270 classrooms.

Sample attrition also significantly impacts the statistical power available in the data

analysis.

I chose a target of 100 classrooms (MDE of 0.25) for this study, but as I note

below only 50 classrooms could be recruited. The MDE of 50 classrooms is 0.35 and

approximately 0.31 or smaller depending on the explanatory strength of covariates. I also

ran post-hoc power calculations using new intra-class correlations for each dependent

variable (DV). One DV yielded an ICC of 0.15 (as in Table 4.1). Three dependent

variables produced an ICC between 0.01 or 0.02, which indicates that most of the

variation occurs between individuals and not clusters. Table 4.2 shows the required

sample sizes at these ICC levels by minimal detectable effect (holding all other

assumptions constant).

Table 4.2: Required Classroom Clusters by ICC x MDE

Intra-Class Correlation
Minimal Detectable Effect 0.01 0.02
0.10 150 180
0.20 40 45
0.30 17 20
0.40 10 12
 100

Clearly, when clustering is minimal one needs fewer classrooms to detect smaller

effects. As Table 4.2 indicates, a sample of 50 classrooms yields sufficient power to

detect effects as small as 0.20 standard deviations when clustering is minimal. When

cluster-effects are present (i.e. an ICC = 0.15 in Table 4.1) a sample of 50 classrooms

may detect a MDE of 0.25-0.35 standard deviation changes in the best case.

Sample and Randomization

With the assistance of the 2 districts, teachers were identified that were

particularly known for their motivation and interest in using technology in their

classrooms. In education reform, a common obstacle is local implementation

(McLaughlin, 1990). K-12 technology researchers frequently observe that although

classrooms receive much investment in computer technology, teachers infrequently or

almost never adopt the tools in practice (Cuban, 2001). With little or no use, one cannot

expect any effects of a new technology on educational outcomes. Thus, implementation

looms large in any new initiative or intervention. I attempted to control for this major

factor by recruiting teachers who were most likely to utilize a new technology. District

leaders and high school principles nominated teachers who fit this profile and they were

sent an email asking for their participation. With the help of the district technology

coordinators, I also visited the high schools to speak in-person with the nominated

teachers.

A total of 10 high school teachers agreed to participate in the study: 5 Social

Studies, 4 English, and 1 Science. The teachers taught an average of 5 class periods,

which resulted in a sample of 50 separate class periods. Periods had an average class size
 101

of 27 students with a total sample of 1,349. To control for any effects of teacher or

subject matter, class periods were randomized within each teacher. For example, each of

the 10 teachers had class periods in both the treatment and control groups. After

randomization, 22 class periods and 528 students were assigned to the experimental SNS

site. In addition, 28 class periods and 821 students were assigned to the control group.

Descriptive statistics of the classrooms are provided in Table 4.3 below.

Table 4.3: Characteristics of Classrooms in SNS and Control Groups

Social Network Site Control Group

(n = 22) (n = 28)

% Male 44% 51%

% Female 56% 49%

% White 42% 34%

% Hispanic 30% 37%
% Black 4% 3%
% Asian 23% 26%
% Other < 1% < 1%

% Free Lunch 21% 17%

Prior GPA 3.0 2.9

ELA CST Score 374 368

ELA CST = English Language Arts, California Standards Test

Classrooms in the treatment and control groups did not differ significantly along

demographic indicators. The only statistically significant difference occurred with

gender. The control group had a higher proportion of male students compared to the SNS

group. The classrooms did not differ in measures of prior achievement including grade

point average (GPA) and standardized test scores. The results suggest that the

randomization procedure worked fairly well for this sample of 50 class periods.
 102

One major challenge of this study is the possible cross-contamination across

treatment and control groups. Many randomized studies of classrooms or students face

this potential obstacle because students in the same school may likely know who has been

placed in either group. The potential exists for students to speak to one another or behave

in ways that reduce the effectiveness of the intervention. In this study, class periods were

the unit of randomization. Thus, it is possible for students to be placed in both the

experimental or control groups if they were members of specific class periods. In the total

sample of 1,349 observations there were 111 students who are duplicated or were

members of multiple classrooms. Of 111 students, 45 received both the SNS intervention

and control based on their classroom membership. This minimal level of cross-

contamination does not pose a large threat to the study design. I collected various

measures of how the SNS was used, also known as implementation fidelity (detailed

below). The overall level of student and teacher use of the SNS was extremely low, and

the threat of treatment and control spillover is quite minimal.

Social Network Site Intervention

A private social network site was built for use in the 2 districts. The network site

approximated all of the functionality seen on popular SNS. Students and teachers could

create profiles, add friends to their social networks, and post a variety of data including

public text messages (often called wall posts), comments, personal messages, pictures,

video, and other files. The site also allowed for blogging and wiki capabilities. Blogs are
 103

personal journals that are written online and can be viewed by other users. Wikis are

collaborative pages that multiple users can edit together to create shared information

resources.

At the beginning of the 2009-2010 school year I met individually with each

teacher. During this meeting, teachers were trained on how to utilize the website and all

research procedures including timeline and survey distribution. In addition, I worked with

each teacher to brainstorm ways to utilize the SNS for their particular courses and topic

areas. For example, two English teachers were teaching a module on the Scarlett Letter

and planned to have their students upload their writing samples to the SNS. They planned

to encourage their students to comment and edit their peers’ assignments over the time

span. The science teacher planned to use the SNS for students to collaboratively create

Wiki pages based on their Earth Science and Physics concepts. His hope was for students

to create study guides together that reviewed the topics they covered. The goal of the

individual teacher meetings was to plan how each teacher might implement the SNS as it

made sense for their particular curriculum and practice. Immediately after the teacher

meetings and orientation, the SNS classrooms used the site for a period of 6 weeks.

Data Collection

At the end of the evaluation period, teachers implemented the posttest survey. The

survey (Appendix A) included questions that assessed students’ level of school

engagement, offline social capital and online social capital (discussed in detail below).

Students were asked if they were Facebook or MySpace users. Students and teachers

were also given open-ended response questions that asked how often they used the SNS
 104

and reasons for their use or non-use of the intervention. The posttest response yielded

complete data for approximately 870-930 students (depending on the various outcome

measures). In addition to survey data, I collected information on website usage. The

social network site is built with an underlying database that records user activity. The

database provides information such as the number log-ins and interactions that occurred

on the website during the 6 week period. This quantitative data combined with the

qualitative survey responses supply evidence to the fidelity of implementation, or how

much the teachers and students actually used the tool.

Finally, the participating school districts provided administrative data on all

students who participated in the study. I obtained demographic information including

age, gender, ethnicity, and free lunch status. In addition, the districts provided two

measures of prior academic achievement: the students’ cumulative GPA prior to the

beginning of the school year and their English test scores on the California Standards

Test CST). I merged this administrative data with the survey responses to form the

complete dataset.

Outcome Measures and Covariates

Hypotheses 1-4 consider the effects of social network sites on students’ social

capital, school engagement, and academic achievement. I collected 4 outcome measures

to examine these confirmatory questions.

School Social Capital (SSC): To assess students’ level of school social capital, I

utilized the Internet Social Capital Scales (ISCS) from Williams (2006). The scale was

modified from questions that measure an individual’s offline social capital. Four
 105

questions create the summated scale and assess the quality of students’ relationships in

their school community (see Appendix A, Question 8a-d). Students were asked to what

degree they agreed with various statements concerning social support in their school

(range from 1-4). In terms of internal reliability, Cronbach’s α for this scale was .715.

Online Social Capital (OSC): The Internet Social Capital Scales also consist of

items to measure individuals’ level of social capital in online relationships (Williams,

2006). Five items were used to ask students how connected they felt to others based on

their online interactions (Appendix, A, Question 8f-j). Like the SSC scale, students

responded with levels of agreement to each question (range from 1-4). The Cronbach’s α

for this scale was .909.

School Engagement (SE): In this study, I utilize a modified measure of

emotional engagement with school developed by Christophel (1990). The scale consists

of 10 questions that ask students how they feel about their school (Appendix A, Question

6). The original scale asked students about a particular class, but was modified in this

study to ask students about their school community in general. Students indicate from a

1-5 scale their feelings about high school; for example how excited or unexcited they feel

about their school. The Cronbach’s α for the school engagement scale was .903.

Course Performance (GPA): The final outcome measure was an indicator of

students’ academic performance. Teachers provided estimates of student course grades at

the end of the experiment, which were converted into a 0-4 GPA scale.

Covariates: As noted earlier, the power calculations show increased statistical

power when one utilizes regression models with baseline covariates. To maximize
 106

statistical power given the sample of 50 classrooms, I utilize two control variables in the

data analysis. The models include students prior cumulative GPA and gender. I include a

measure of prior academic achievement because it is a strong predictor of future

achievement and also correlated to social capital and engagement (Dika & Singh, 2002;

Fredericks et al., 2004). I include gender because the experimental and control groups

still differed in gender makeup after randomization. The reference group for gender is

female.

In addition to the control variables, I also include three variables that directly

relate to the exploratory question in this study. Three dummy variables describe which

social network sites a student is a member of: only Facebook, only MySpace, or whether

the student uses both. The reference group is students that do not use any SNS. The

exploratory research question asks whether student use of already existing networks has

any relationship to social capital, school engagement, or achievement. These variables

allow an analysis of the relationship in comparison to the effects of the experimental

network site.

Fidelity of Intervention Measures: To assess the fidelity of treatment I utilize

two data sources. First, open-ended survey responses from teachers and students offer

evidence of how well the classrooms utilized the social network site. The qualitative data

also sheds some light on the challenges and potential of SNS in educational contexts. In

addition, I coded the website database, which includes a record of all users and their
 107

activity during the experiment. The data includes how many teachers and students logged

into the site, the number of times they updated their profile, and the number of times they

interacted with others on the SNS.

Table 4.4: Descriptive Statistics of Dependent Variables and Covariates

Mean Standard Deviation Min Max

School Social Capital 12.29 2.42 4 16
Online Social Capital 13.07 4.03 5 20
School Engagement 34.47 8.13 10 50
Course Grade 2.68 1.33 0 4

Prior GPA 2.96 0.87 0 4

Gender 0.49 0.49 0 1
Facebook 0.21 0.41 0 1
MySpace 0.26 0.44 0 1
Both SNS 0.35 0.48 0 1

Data Analysis

I utilize a random effects (RE) model that accounts for the classroom-clustered

data in the research design. The model is specified as:

Yij = β0j + β1treatmentj + β2facebookij + β3myspaceij + β4bothsnsij +

β5prior_gpaij + β6genderij + uj + rij

In the model, Yij refers to the dependent variable for student i in classroom j. The

intercept β0j is the population average of classrooms and the term uj represents the

between classroom variance. The coefficient β1 is the effect of the experimental treatment

and is the primary focus of hypotheses 1-4. The coefficients β2 - β4 are the relationships

between students’ existing social networks and the outcome. These variables – facebook,

myspace, and bothsns – directly address the exploratory research question of whether
 108

existing social networks are related to students’ outcomes. The model includes control

variables for prior achievement, gender, and finally the term rij represents the within

classroom variance. Note that the RE regression is one case of a multi-level or

hierarchical linear model (HLM), where the intercept or classrooms (β0j + uj) are modeled

as random effects.

Missing Data

Missing data on the outcome variables was a significant challenge. The majority

of district data such as student demographics and prior achievement had nearly no

missing data. However, outcome variables had significant missing values. Table 4.5

summarizes the frequencies of subjects with complete and missing data based on the

dependent variables in this analysis. For every model, approximately 30-35% of data

consisted of missing data. The key question is whether the data is missing at random

(MAR), as then one might utilize imputation methods for the missing values (Royston,

2004). Examining the correlations between missing-ness on each outcome and student

background variables showed small relationships. Most correlations ranged from 0.00 to

0.15 and suggest that data were not missing completely at random (MCAR).

Nevertheless, the very small correlations between missing-ness and covariates are

evidence for MAR.

 109

Table 4.5: Percentage of Missing Data on Dependent Variables

Dependent Variable Complete (%) Missing (%) Total

Course Grade 936 (69%) 413 (31%) 1,349

School Social Capital 869 (64%) 480 (36%) 1,349

Online Social Capital 873 (65%) 476 (35%) 1,349

School Engagement 880 (65%) 469 (35%) 1,349

With nearly 35% of missing outcome data, the effective cluster sample size

ranges from 35-41 classrooms. This reduction severely impacts statistical power and

since missing-ness was not MCAR, might bias beta estimates and standard errors. To

alleviate these issues, under the assumption that data was MAR, I utilized multiple

imputation using the ice program available in STATA (Royston, 2004). The algorithm

creates 5 datasets of imputed values based on existing covariates. Regression analysis is

conducted on each individual dataset, and the final beta and standard error estimates are

derived from the average of these separate analyses. I compared results using both list-

wise deletion (no missing data handling) and the imputed datasets. The general trend with

the imputed data shows no significant differences in beta estimates. The standard errors

were slightly higher in the imputed data leading to somewhat more conservative tests of

significance. However, these differences did not change any interpretations or

conclusions from the data.

Findings

Fidelity of Intervention

The overall implementation of the social network site faced numerous obstacles.

Only 203 of the 528 students in the treatment group (38%) used the experimental SNS.
 110

Their activity also ranged dramatically. On average, the participating students logged in 4

times in the 6-week period. They updated their profiles an average of 2 times. They

interacted in the website – created content, wrote messages, or shared files – an average

of 42 times. This activity level pales in comparison to the hours youth spend on their

already existing networks on Facebook and MySpace. Logins, profile updates, and

interactions would plausibly reach thousands in a similar 6-week period.

While all efforts were made to recruit teachers that were comfortable with

technology and motivated to participate in the study, the general level of teacher adoption

was low. Only 5 of the 10 teachers logged into the website during the 6-week period. Of

these individuals, 3 teachers logged in more than once to create activities for their

students. In follow-up responses, the teachers identified numerous reasons for the

difficulty in using the SNS. Most all of the teachers cited lack of available technology in

their schools and student motivation as the main factors in non-use. For example, an

English teacher stated, “The greatest challenge was computer availability. Some students

also struggled with the log in process. What also made the assignment process difficult

was that only about half the kids were doing it per period so it was difficult to grade.”

Several teachers noted that it is difficult to motivate students to participate in

something new, when overall levels of motivation were quite low for academics. One

Social Studies teacher stated, “my first period class – most do nothing so I knew they

would not get into this…” and another English teacher observed that, “The potential is

there; it could be a very useful tool, but I did not get my students to buy in, yet.” The

most pervasive comment by teachers spoke to the difficulty of adding new curriculum or
 111

projects to their existing plans. One teacher’s observations communicated this general

sentiment:

The biggest problem was that the technology was tacked on to an established
curriculum and routine. The students saw that I was testing the waters (I was
quite up-front about this) and that failure to participate would not have more than
a minor effect on their grades. For many students, the grade is god; if I don't
lower a grade for failure to complete work, they will not do the work. I imagine
that if I set up my profile and got a few groups, links, blog posts, and other "stuff"
on the site before the first day of school, the tool would be more useful.

As with many interventions and reforms, the organizational and cultural structure of

public schools provided significant obstacles for teachers to implement a new technology.

Not surprisingly, student response to the SNS was indifferent at best. Many

students logged in, created profiles, and chatted with others on the site. A few students

uploaded assignments or examples of their work, but academic activity was very

minimal. The majority of students indicated that the SNS was not particularly useful for

them because it did not relate to their final course grades.

However, many students felt that social network sites could be beneficial for

school-based goals. Several students suggested that they would definitely use SNS for

clubs, activities, and to keep in touch with everything happening on campus. Some

students noted that, “People feel comfortable with expressing themselves as well” and

viewed SNS as an outlet for student communication. Many youth felt that students might

use SNS to help each other to complete homework and assignments. Finally, many

participants noted that their peers already use other social network sites. When asked if

SNS should be used in school, one teenager stated, “They already are, just not directly…
 112

just to know about homework or a test.” Another student also raises the possibility for

schools to leverage their already existing networks on sites such as Facebook and

MySpace:

I think the problem is that there are already too many social networking sites. This
kind of idea needs to be incorporated into existing sites as students are not likely
to switch.

The qualitative responses from teachers and students suggest two considerations. First,

implementing a new technology in school settings requires overcoming many obstacles

including teacher adoption, student motivation, and the structure of curriculum and

assessment in K-12 education. With minimal adoption, particularly in this study, it is

unlikely that technology evaluators will find significant effects on students’ educational

outcomes due to social media tools. Second, as the students noted in their survey

responses, schools might be able to leverage already existing tools and social networks to

engage students. I evaluate these possibilities in the next sections.

Hypothesis 1: Did the Experimental SNS Improve Students’ School Social Capital?

The first hypothesis considers whether the experimental SNS had any effect on

students’ sense of social capital in their school relationships. The variable – Treatment –

in Table 4.6 below shows that the school-based SNS had no statistically significant effect

on school social capital. The coefficient is also negative, which indicates that the

experimental SNS slightly lowered students’ feelings of connection with their school

community. The findings, when coupled with the implementation evidence, suggest the

possibility that social network sites could have potentially negative effects on school

communities when implemented poorly. A SNS is fundamentally dependent on user

 113

interactions and if no one is using the site, the online community can be a lonely place.

Perhaps the minimal adoption of the experimental SNS actually left students feeling less

connected to their peers. One cannot assess whether a successfully implemented SNS

would improve school-based relationships, but the results of this study offer evidence that

how educators implement social technologies in classrooms can potentially have negative

effects if done poorly.

Table 4.6: Result of Random Effects Models for each Dependent Variable

School Social Online Social School Course Grade

Capital Capital Engagement

Intercept 11.21 10.06 29.42 -0.49

(0.46) (0.60) (1.85) (0.20)
Treatment -0.25 -0.22 0.11 -0.26
(0.19) (0.35) (0.72) (0.14)
Facebook -0.07 2.27** -0.003 -0.02
(0.34) (0.41) (0.84) (0.09)
MySpace 0.09 1.51** 0.53 0.02
(0.36) (0.37) (0.81) (0.12)
Both SNS 0.47 3.05** 0.61 0.04
(0.27) (0.33) (0.65) (0.07)
Prior 0.37** 0.35** 1.65** 1.14**
Cumulative (0.09) (0.15) (0.44) (0.04)
GPA
Gender -0.16 0.24 -0.47 -0.06
(0.25) (0.31) (0.64) (0.07)

Standard errors in parentheses, * p < 0.10, ** p < 0.05

Hypothesis 2: Did the Experimental SNS Improve Students’ Online Social Capital?

The second hypothesis examines whether using the experimental social network

site improves students’ social capital in their online relationships. The Treatment variable

(Table 4.6) did not have a statistically significant effect on students’ Online Social

Capital. Similar to the case of School Social Capital, the negligible use of the intervention

probably offered few opportunities for youth to develop relationships in the online
 114

community. When 70% of your peers are not using the site you are on, the online

community may feel like an isolating experience.

Hypothesis 3: Did the Experimental SNS Improve Students’ Emotional Engagement with

School?

The third hypothesis considered whether using the school SNS improved

students’ emotional engagement with school. The Treatment variable (Table 4.6) did not

have a statistically significant effect on how engaged students felt with school. Similar to

hypotheses 1 and 2, the evidence shows that a school-based SNS site will have no effect

on students’ emotional engagement with school when the technology is not received well

by teachers and students. The finding appears to contradict the hope of the majority of

education administrators, who believe social media might help improve student

engagement. However, this study only shows that such gains will not occur from poorly

implemented technology alone. The question, “What would happen if we successfully

connected students via social technologies in school?” still remains for investigation.

Perhaps a vibrant and active online school community may yet increase student

engagement.

Hypothesis 4: Did the Experimental SNS Improve Student Achievement?

Finally, the experimental SNS had no significant effect on students’ course grades

(Table 4.6, Treatment variable). The single greatest predictor of a students’ performance

in a given class was their cumulative GPA prior to the start of the school year. Not

surprisingly, a student’s prior academic history largely determined how well they

performed in any given class or subject. Again, the minimal use of the social network site
 115

prevents any definitive conclusions from this analysis. The results offer practical

implications however. Merely introducing social technologies into formal classrooms do

not hold much promise to improve students’ academic performance. As many educational

media scholars have noted in past research, the pedagogies that teachers employ with the

social media tool may offer the most plausible connections to improved student learning.

Exploratory Results: School SNS vs. Facebook and MySpace

The exploratory research question asks whether a school-based network site has

differential effects on students compared to their already existing networks in Facebook

and MySpace. The variables Facebook, MySpace, and Both SNS (Table 4.6) directly

allow for a comparison of different online communities with the experimental treatment.

For School Social Capital, none of the social network sites appear to affect students’ level

of connection to their school community. However, those students who used both

Facebook and MySpace had a 0.47 increase in their social capital score compared to

students who do not use any SNS (α < 0.104). This beta estimate yields an effect size of

approximately 0.19 standard deviations which would reach significance had this study

been conducted with a larger sample of classrooms (i.e. 100). The results suggest, albeit

exploratory, that those students who are particularly adept at developing relationships in

social network sites could be better connected in their school communities.

The relationship between SNS use and Online Social Capital is even more

striking. The school-based SNS had no significant effect. However, students who use

Facebook had substantially higher online social capital than their peers who do not use

any SNS (see Table 4.6). Facebook users scored approximately 0.56 standard deviations
 116

higher in Online Social Capital. MySpace users also had significantly higher social

capital, at roughly 0.37 standard deviations. Finally, students who are members of both

Facebook and MySpace scored about 0.75 standard deviations higher in social capital

than peers who do not use any. These teenagers are what scholars might call super-

communicators, who use social technology and media tools to connect with wider

networks (Lenhart, Madden, Macgill, & Smith, 2007b). It appears that high school youth

develop positive relationships and social capital in their online activities.

These exploratory findings provide evidence for a critical question facing K-12

administrators and technology directors: Should schools ban access to tools like

Facebook and MySpace, and instead use district-controlled technology? The majority of

school districts utilize this strategy because of concerns about student safety in these

online communities. However, this study demonstrates that a school-based social network

site produces no significant benefits, and at times slightly negative effects on students. A

district-imposed SNS that is poorly used and frankly “un-cool” for students may not

produce the benefits of strengthening student relationships and engagement. Instead, as

the students noted in open-ended survey responses, youth are already members of

existing social networks sites like Facebook and MySpace. These teenagers are also

developing relationships and bonds, social capital, in these communities. If the goal is to

strengthen student connections to their school, the best strategy may be to meet students

where they already are, instead of imposing unnecessary constraints.

 117

Chapter 4 Conclusions

Technology is here to stay. K-12 educational institutions may be the last major

sector to fully utilize new digital tools (Cuban, 2001). However, students are increasingly

wired and connected via computers, mobile devices, and social media tools like SNS. The

pressure for schools to use social media to engage students is increasing as policymakers,

researchers, and adults begin to comprehend how integrated these tools are in youth life

(Jenkins, 2006). This study offers an evaluation of a popular tool – the social network site

– and its effect on student outcomes. The results of this study suggest that a school-

imposed social network, which is poorly received by students, will have no significant

effect and perhaps negative effects on student-school relationships, engagement, and

GPA. However, students’ existing social networks on sites such as Facebook and

MySpace may provide ways to improve student relationships with peers, teachers, and

the school community at large. Perhaps K-12 educators might consider leveraging these

existing social networks, rather than banning access to these technologies as many

districts currently do.

The cursory cost-benefit implications of these findings are substantial. Tools such

as Facebook and MySpace are free and widely available. The main costs for school

districts arise from the need to develop policies that allow student access to these sites,

while avoiding issues of youth safety and inappropriate access to online material. These

costs seem quite minimal when compared to the benefits of increasing digital literacy

among youth and improving student-school relationships. Just as corporations and

businesses utilize social networking tools to connect with consumers and build brand
 118

loyalty, schools might leverage existing social networks to strengthen relationship

between students and the larger community. The upfront costs are minimal, but the

benefits could be significant.

The limitations of this study also illuminate the challenges of implementing new

technology in K-12 schools, and the subsequent obstacles to realizing positive student

outcomes. The experimental SNS had no significant effect on student outcomes, with

slightly negative relationships to social capital and student achievement. However, much

of these findings can be attributed to the lack of implementation and usage of the site.

Students use Facebook and MySpace with enthusiasm. They did not use the school-based

SNS and felt it was not very hip or useful. When critiquing this study on grounds of

research design, one would note this behavior as lack of implementation fidelity. In

actual K-12 schools, these trends are real challenges to overcome in order to realize any

potential effects of social media. Future studies should critically consider

implementation: the training of teachers, extensive planning and integration of the new

technology into existing curriculum, and resources such as coaches to ensure that

teachers in fact implement the intervention. Similarly, school districts might think twice

about building and utilizing their own in-house social network tools. A poorly

implemented site will have no benefits. However, students are already invested in

existing sites such as Facebook and MySpace. The more effective option may be to

leverage this energy in positive ways, instead of imposing obstacles.

Another limitation of this study concerns the sample size of 50 classrooms. Future

studies will need to consider larger samples, nearing 100 classrooms or more, to have
 119

sufficient power to detect smaller (but substantial) effects. The issue of randomized

control also plays a contradictory role when evaluating social media. Randomized control

allows the researcher to more effectively attribute cause to the intervention. However,

when randomizing students or classrooms to use a social network, one necessarily

delimits the social network.

In this study, students in the treatment classrooms could have had friends in the

control classrooms. Since some friends were not included in the network site, treatment

students might not be as compelled to use the site since their friends were not using the

site. Social technology relies on connection and interaction with networks. Randomized

control relies on separation into distinct comparison groups. Future studies may have to

consider entire schools as the unit of randomization, to improve implementation and

realize any network effects. These designs will require substantially more financial and

human resources.

The most promising future studies might consider utilizing Facebook, MySpace,

or other social media tools. Researchers might create school-based applications (i.e. apps)

or group pages on these sites. One cannot randomly assign students to Facebook or

MySpace, but participants might be randomly invited to use the app or join the network

group. These types of interventions might ensure higher implementation fidelity, while

offering better external validity as to how these technologies affect students, schools, and

education communities. This study is one of the first to (a) examine the media effects of

social technology on youth, (b) consider outcomes that relate to education concerns, and
 120

(c) utilize a randomized control design. Future studies in this vein will be imperative as

technology becomes ever more present in the way educators deliver education, the way

youth learn, and the way society communicates in the future.

 121

Chapter 5: Future Directions for Research on Social Network Sites

The chapters in this dissertation contribute several insights into the use of social

network sites amongst teenage youth. A review of the extant literature (chapter 2)

suggests that SNS research is currently in a very early, exploratory stage of development.

The majority of studies focus on defining social network sites themselves, or examining

the cultural and social practices of particular online communities such as Facebook,

MySpace, YouTube, or Dodgeball (boyd, 2006; boyd & Ellison, 2007; Humphries, 2007;

Walther et al., 2008; Walther et al., 2009). Researchers are also interested in usage

patterns, or whether there are gaps in access to these new social media tools (Hargittai,

2007; Lenhart et al., 2007b). Even less evidence exists surrounding the use and effect of

social technologies on youth, their development or their academic success (Ito et al.,

2009; Karpinski, 2009a).

The two empirical studies presented here offer additional contributions to this

nascent field of research. Chapter 3 examined digital divides among teenage populations

and their access to SNS, and was a direct extension of past digital divide research in adult

populations (i.e. Hargittai, 2007). The findings suggest that teenage youth are

increasingly connected to social network sites and that these online communities may

play an increasing role in their daily interactions and communication. Chapter 4

presented one of the first attempts to experimentally discern any social and learning

effects of SNS on teenage youth. The cluster-randomized trial findings teach us that

social network sites, as a technological tool, do not significantly affect youth social
 122

capital, psychological well-being, or their academic achievement. However, I find

exploratory evidence that teenagers’ existing social networks on Facebook and MySpace

were significantly, and positively related to their levels of social capital. The studies offer

two major implications for future research in this domain: (a) the need to develop a

coherent research stream amidst a sea of new social media tools, and (b) the importance

of taking into account social context when studying online communities.

Making Sense of Social Network Sites as They Evolve

Studies of social network sites will likely become a major area of focus in the

coming years. New SNS communities surface everyday. For example, early sites such as

Friendster and MySpace gave way to Facebook, and only recently has Twitter become a

highly utilized social network site. Current research typically focuses on particular online

communities. Some sites, such as Facebook, have succeeded and others, such as

Dodgeball, have disappeared. A multitude of new social network sites emerge and fade

away each day. One critical challenge for future SNS researchers will be to make sense of

the different communities over time and create a coherent picture of socially networked

communication. Studies of particular communities are liable to become quickly obsolete

as technology evolves rapidly.

However, SNS researchers have a unique opportunity to create coherent findings

because social network sites have quickly become the over-riding paradigm of Internet

technologies. As Livingstone (2008) observes, SNS appear in just about every website

available today. For example, YouTube is a video sharing site but is structured around

profiles and friend networks. One can now visit CNN.com to read the daily news, but
 123

also share articles by linking their Facebook accounts. Even traditional Internet services

such as email integrate social network site features, such as the Gmail service from

Google. The definition of SNS – profiles, friends, and networks – remains stable, but its

implementation across new Internet technologies is varied. This situation affords future

researchers the ability to work from a similar foundation, but also critically examine how

social network sites are used in different contexts.

There will always be renewed interest in whichever new technology emerges in a

given time period. Research that closely examines and documents new media tools, the

populations that use them, and the cultural conditions that arise from them, will always be

vital. However, scholars will make the most impact if they can begin to relate new social

technologies to past media tools. How does a given technology’s features and design

differ, or align, with other major media platforms? What are the populations that use a

given technology, and how are they similar or different than users of other social network

sites? What are the cultural and social norms that govern interaction and communication

in different social network sites? How are these communication patterns related to the

design and function of the online community? These are the type of questions that can

simultaneously describe new developments in technology, but also link to past research

to create a coherent field of study.

The Importance of the Social Context Surrounding SNS

Another implication of this dissertation is the need for future research to consider

the social context that surrounds a given technology. The randomized trial presented in

Chapter 4 highlights how a strict media effects framework ultimately offers limited
 124

insight into the impact of social network sites on youth. The treatment in this experiment

was a social network site that was introduced into a high school classroom context.

However, the lack of use of the technology is likely the main contributor to the non-

significant effects found for the students. The treatment, or lack of use of the SNS,

resulted in no significant effects on high school youth on any of the outcomes: social,

psychological, or academic. As previous media researchers have found, a technology

itself has no direct effect on social outcomes because it is confounded with the use of the

tool (Clark, 1983; Clark, 1991; Kling, 2007).

Scholars from a variety of fields agree that new technologies interact with social

factors to ultimately determine how a tool is used, and what effects they have on

organizations and individuals (Fulk & DeSanctis, 1999; DeSanctis & Poole, 1994; Kling,

2007). A given technology can be utilized in a variety of ways depending on the user,

their goals, and their social context. Social network sites are also utilized in diverse ways.

Youth use them to socialize with friends. Corporations utilize social media to advertise

their products and build customer loyalty. Professionals use them to keep abreast of new

information. Political campaigns utilize social media to mobilize voters. Schools might

use them to network students and enhance academic activities.

Given the disparate uses of the same technology, it is difficult to attribute any

causal effects to a social network site. A SNS alone does not cause youth to learn more or

less effectively. However, teachers might use a social network technology in such a way

that leads to highly effective student learning. The onus for future education technology

researchers will be to discern: (a) how new social media tools change the teaching and
 125

learning process, (b) what educational practices lead to improved student learning, (c) for

which populations these practices are most effective, and (d) how the cultural and

institutional setting of formal and informal education affects the use and subsequent

effects of a technology. The technology, setting, audience, and utilization are all factors

that future researchers must control for when examining the media effects of social

network sites.

Chapter 5 Conclusions

Teenagers spend a considerable amount of time in social network sites. These

online communities represent a significant setting – perhaps just as noteworthy as home

and school – within which youth socialize, learn, and develop. The current moment in

history affords scholars a unique opportunity to examine SNS, with the adoption of the

technology skyrocketing in just the past few years. The studies in this dissertation

contribute to the nascent discussion surrounding social network sites and teenagers, but

much work remains in this field. As new social technologies emerge and more youth

become connected to online networks, questions of who accesses these communities and

what effects they have on youth become vital considerations for parents, educators, and

policymakers.

Sharples et al. (2009) observe that, “At present, schools are caught between the

rock of parental fears about Internet abuse and the hard place of helping children to

develop responsible and creative use of Web 2.0 for learning” (p. 82). Researchers

interested of SNS have a unique opportunity to inform these concerns. Continued

research in this area is needed to inform critical education policy concerns. Should
 126

schools block access to SNS and other social media? How should teaching and learning

occur with social tools like SNS? Finally, Zhao & Lei (2009) note that, “today’s children

are spending more and more time in the digital world… Future research needs to explore

the psychological and cognitive impacts of technology use on children, to understand

how time spent on the Internet affect children developmentally” (p. 688). Technology

will never go away, and social media will continue to shape our children’s life

experiences in ever-deeper ways. The question of how new technologies impact our next

generation of youth, both socially and intellectually, becomes one of the most pivotal

questions for both scholars and society as a whole.

 127

Bibliography

Alexa. (n.d.). Global top sites. Retrieved on May 21, 2009 from
http://www.alexa.com/site/ds/top_sites.

Aslanidou, S., & Menexis, G. (2008). Youth and the internet: Uses and practices in the
home. Computers & Education, 51(3), 1375-1391.

Bandura, A. (2002). Social cognitive theory of mass communication. In J. Bryant & D.

Zillman (Eds.), Media Effects: Advances in Theory and Research (pp. 121-153).
Mahwah, NJ: Lawrence Erlbaum Associates.

Barton, P. E. (2005). One-third of a nation: Rising dropout rates and declining

opportunity. Princeton, NJ: Policy Information Center, Educational Testing
Service.

Beaudoin, C. E. (2008). Explaining the relationship between internet use and

interpersonal trust: Taking into account motivation and information overload.
Journal of Computer-Mediated Communication, 13, 550-568.

Berliner, D. C. (2002). Educational research: The hardest science of all. Educational

Researcher, 31(8), 18-20.

Bernard, R. M., Abrami, P. C., Lou, Y., Borokhovski, E., Wade, A., Wozney, L., et al.
(2004). How does distance education compare with classroom instruction? A
meta-analysis of the empirical literature. Review of Educational Research, 74(3),
379-439.

Blanding, M. (2009, Winter). Thanks for the add: Now help me with my homework. Ed.
Magazine: The Magazine of the Harvard Graduate School of Education, 29-33.

Bourdieu, P. (1986). The forms of capital. In J.G. Richardson (Ed.), Handbook of theory
and research for the sociology of education (pp. 241-258). New York:
Greenwood Press.

boyd, d. (2006). Friends, friendsters, and myspace top 8: Writing community into being
on social network sites. First Monday, 11(12).
http://www.firstmonday.org/issues/issue11_12/boyd/index.html.

boyd, d. m. (2008). Why youth (heart) social network sites: The role of networked
publics in teenage social life. In D. Buckingham (Ed.), Youth, Identity, and
Digital Media (pp. 119-142). The John D. and Catherine T. MacArthur
Foundation Series on Digital Media and Learning. Cambridge, MA: The MIT
Press.
 128

boyd, d. m. (2009). It is easy to fall in love with technology… Retrieved on February 21,
2010 from http://wp.nmc.org/future/ideas/danah-boyd/.

boyd, d. m. (n.d.). Friendship. In M. Ito, H. A. Horst, M. Bittanti, d. boyd, B. Herr-

Stephenson, P. G. Lange et al. (Eds.), Hanging around, messing around, and
geeking out: Living and learning with new media. Cambridge, MA: MIT Press.
http://digitalyouth.ischool.berkeley.edu/book-friendship.

boyd, d. m. & Ellison, N. B. (2007). Social network sites: Definition, history, and
scholarship. Journal of Computer-Mediated Communication, 13(1), article 11.

Burkhardt, H. & Schoenfeld, A. (2003). Improving educational research: Toward a more

useful, more influential and better-funded enterprise. Educational Researcher,
31(9), 3-14.

Cataldi, E. F., Laird, J. & KewalRamani, A. (2009). High school dropout and completion
rates in the United States: 2007. Washington D.C.: National Center for Education
Statistics, Institute of Education Sciences, U.S. Department of Education.

Cheong, P. H. (2008). The young and techless?: Investigating internet use and problem-
solving behaviors of young adults in Singapore. New Media & Society, 10(5),
771-791.

Christophel, D. M. (1990). The relationships among teacher immediacy behaviors,

student motivation, and learning. Communication Education, 39, 323-340.

Clark, R. E. (1983). Reconsidering research on learning from media. Review of

Educational Research, 53(4), 445-459.

Clark, R. E. (1991). When researchers swim upstream: Reflections on an unpopular

argument about learning from media. Educational Technology, 31(2), 34-40.

Clark, R. E., Yates, K., Early, S., & Moulton, K. (In Press). An analysis of the failure of
electronic media and discovery-based learning: Evidence for the performance
benefits of guided training methods. In Silber, K. H. & Foshay, R. (Eds.),
Handbook of Training and Improving Workplace Performance, Volume I:
Instructional Design and Training Delivery. Somerset, NJ: Wiley.

Coleman, J. (1990). Foundations of social theory. Cambridge, MA: Harvard University

Press.

Cuban, L. (2001). Oversold and underused: Computers in the classroom. Cambridge,

MA: Harvard University Press.
 129

DeSanctis, G., & Poole, M. S. (1994). Capturing the complexity in advanced technology
use: Adaptive structuration theory. Organization Science, 5(2), 121-147.

Dika, S. L., & Singh, K. (2002). Applications of social capital in educational literature: A
critical synthesis. Review of Educational Research, 72(1), 31-60.

Donath, J. (2007). Signals in social supernets. Journal of Computer-Mediated

Communication, 13(1), article 12.
http://jcmc.indiana.edu/vol13/issue1/donath.html.

Donath, J., & boyd, d. (2004). Public displays of connection. BT Technology Journal,
22(4), 71-82.

Donner, A., & Klar, N. (2000). Design and analysis of cluster randomization trials in
health research. New York: Oxford University Press.

Dwyer, C., Hiltz, S. R., & Passerini, K. (2007). Trust and privacy concern within social
networking sites: A comparison of facebook and myspace. Paper presented at the
Thirteenth Americas Conference on Information Systems, Keystone, CO.

Eccles, J. S., & Templeton, J. (2002). Extracurricular and other after-school activities for
youth. Review of Research in Education, 26, 113-180.

Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of facebook “friends”:
Social capital and college students’ use of online social network sites. Journal of
Computer-Mediated Communication, 12(4), article 1.
http://jcmc.indiana.edu/vol12/issue4/ellison.html.

Emmers-Sommer, T. M., & Allen, M. (1999). Surveying the effect of media effects: A
meta-analytic summary of the media effects research in human communication
research. Human Communication Research, 25(4), 478-497.

Eveland, W. P. (2003). A “mix of attributes” approach to the study of media effects and
new communication technologies. Journal of Communication, 53(3), 395-410.

Facebook. (n.d.). Statistics. Retrieved on April 19, 2010 from

http://www.facebook.com/press/info.php?statistics.

Federal Communications Commission. (n.d.). E-rate. Retrieved on December 4, 2008

from http://www.fcc.gov/learnnet/.

Federal Communications Commission. (n.d.). Children’s Internet Protection Act.

Retrieved on May 28, 2009 from
http://www.fcc.gov/cgb/consumerfacts/cipa.html.
 130

Feldman, A. F., & Matjasko, J. L. (2005). The role of school-based extracurricular

activities in adolescent development: A comprehensive review and future
directions. Review of Educational Research, 75(2), 159-210.

Feuer, M. J., Towne, L., & Shavelson, R. J. (2002). Scientific culture and educational
research. Educational Researcher, 31(8), 4-14.

Finn, J. D., & Achilles, C. M. (1999). Tennessee’s class size study: Findings,
implications, misconceptions. Educational Evaluation and Policy Analysis, 21(2),
97-109.

Fredericks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential
of the concept, state of the evidence. Review of Educational Research, 74(1), 59-
109.

Fulk, J., & DeSanctis, G. (1999). Articulation of communication technology and

organizational form. In J. Fulk & G. Desanctis (Ed.), Shaping Organization
Form: Communication, Connection, and Community (pp. 5-32). Thousand Oaks,
CA: Sage Publications.

Granovetter, M. (1983). The strength of weak ties: A network theory revisited.

Sociological Theory, 1, 201-233.

Greenhow, C., & Robelia, E. (2009). Old communication, new literacies: Social network
sites as social learning resources. Journal of Computer-Mediated Communication,
14(4), 1130-1161.

Greenhow, C., Robelia, B., & Hughes, J. E. (2009). Web 2.0 and classroom research:
What path should we take now? Educational Researcher, 38(4), 246-259.

Gross, E. F., Juvonen, J., & Gable, S. L. (2002). Internet use and well-being in
adolescence. Journal of Social Issues, 58(1), 75-90.

Hare, B. (2009, October 14). Does your social class determine your online social
network? CNN. Retrieved from
http://www.cnn.com/2009/TECH/science/10/13/social.networking.class/index.ht
ml.

Hargittai, E. (2004). Internet access and use in context. New Media & Society, 6(1), 137-
143.

Hargittai, E. (2007). Whose space? Differences among users and non-users of social
network sites. Journal of Computer-Mediated Communication, 13(1), article 14.
http://jcmc.indiana.edu/vol13/issue1/hargittai.html.
 131

Hedges, L. V., & Hedburg, E. C. (2007). Intraclass correlation values for planning group-
randomized trials in education. Educational Evaluation and Policy Analysis,
29(1), 60-87.

Hinduja, S., & Patchin, J. W. (2008). Personal information of adolescents on the internet:
A quantitative content analysis of myspace. Journal of Adolescence, 31, 125-146.

Hohlfeld, T. N., Ritzhaupt, A. D., Barron, A. E., & Kemker, K. (2008). Examining the
digital divide in K-12 public schools: Four-year trends for supporting ICT literacy
in florida. Computers & Education, 51(4), 1648-1663.

Hornik, R. (1981). Out-of-school television and schooling: Hypotheses and methods.

Review of Educational Research, 51(2), 193-214.

Horst, H. A., Herr-Stephenson, B., & Robinson, R. (2009). Media ecologies. In Mizuko
Ito et al. (Eds.), Hanging out, messing around, geeking out: Kids living and
learning with new media. Cambridge, MA: MIT Press.

Hull, G., & Schultz, K. (2001). Literacy and learning out of school: A review of theory
and research. Review of Educational Research, 71(4), 575-611.

Humphries, L. (2007). Mobile social networks and social practice: A case study of
dodgeball. Journal of Computer-Mediated Communication, 13(1), article 17.
http://jcmc.indiana.edu/vol13/issue1/humphreys.html.

Ito, M., Baumer, S., Bittanti, M. boyd, d., & Cody, R. et al., (n.d.). Final report.
Retrieved on April 19, 2010 from http://digitalyouth.ischool.berkeley.edu/report.

Ito, M., Baumer, S., Bittanti, M., boyd, d., Cody, R. et al. (2009). Hanging out, messing
around, and geeking out: Kids living and learning with new media. Cambridge,
MA: The MIT Press.

Jenkins, H. (2006). Confronting the challenges of participatory culture: Media education

for the 21st century. Chicago, IL: The MacArthur Foundation.
http://digitallearning.macfound.org/atf/cf/%7B7E45C7E0-A3E0-4B89-AC9C-
E807E1B0AE4E%7D/JENKINS_WHITE_PAPER.PDF.

Karpinski, A. C. (2009a). A description of facebook use and academic performance

among undergraduate and graduate students. Paper presented at the Annual
Meeting of the American Educational Research Association, San Diego, CA.
 132

Karpinski, A. C. (2009b). A response to reconciling a media sensation with data. First

Monday, 14(5),
http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2503/2183
.

Kennedy, T. L. M., Smith, A., Wells, A. T., & Wellman, B. (2008). Networked families.
Retrieved on November 25, 2008 from
http://www.pewinternet.org/pdfs/PIP_Networked_Family.pdf

Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during
instruction does not work: An analysis of the failure of constructivist, discovery,
problem-based, experiential, and inquiry-based teaching. Educational
Psychologist, 41(2), 75-86.

Kling, R. (2007). What is social informatics and why does it matter? The Information
Society, 23, 205-220.

Kozma, R. B. (1991). Learning with media. Review of Educational Research, 61(2), 179-
211.

Kraut, R., Kiesler, S., Boneva, B., Cummings, J., & Helgeson, V., et al. (2002). Internet
paradox revisited. Journal of Social Issues 58(1), 49-74.

Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., & Mukopadhyay, T., et al. (1998).
The internet paradox: A social technology that reduces social involvement and
psychological well-being? American Psychologist, 53(9), 1017-1031.

Lange, P. G. (2007). Publicly private and privately public: Social networking on youtube.
Journal of Computer-Mediated Communication, 13(1), article 18.
http://jcmc.indiana.edu/vol13/issue1/lange.html.

Lauer, P. A., Akiba, M., Wilkerson, S. B., Apthorp, H. S., & Snow, D., et al. (2006). Out-
of-school-time programs: A meta-analysis of effects for at-risk students. Review
of Educational Research, 76(2), 275-313.

Lemke, C., & Coughlin, E. (2009). Leadership for web 2.0 in education: Promise and
reality. Culver City, CA: Metiri Group. Commissioned by CoSN through support
from the John D. and Catherine T. MacArthur Foundation.

Lenhart, A., Arafeh, S., Smith, A., & Macgill, A. R. (2008). Writing, technology, and
teens. Retrieved on November 25, 2008 from
http://www.pewinternet.org/pdfs/PIP_Writing_Report_FINAL3.pdf.
 133

Lenhart, A., & Madden, M. (2007a). Teens, privacy, & online social networks: How
teens manage their online identities and personal information in the age of
myspace. Washington DC: Pew Internet & American Life Project.

Lenhart, A., Madden, M., Macgill, A. R., & Smith, A. (2007b). Teens and social media.
Washington DC: Pew Internet & American Life Project.

Lewis, K., Kaufman, J., & Christakis, N. (2008). The taste for privacy: An analysis of
college student privacy settings in an online social network. Journal of Computer-
Mediated Communication, 14, 79-100.

Lin, N. (1999). Building a network theory of social capital. Connections, 22, 28-51.

Liu, H. (2007). Social network profiles as taste performances. Journal of Computer-

Mediated Communication, 13(1), article 13.
http://jcmc.indiana.edu/vol13/issue1/liu.html.

Livingstone, S. (2008). Taking risky opportunities in youthful content creation:

Teenagers’ use of social networking site for intimacy, privacy, and self-
expression. New Media & Society, 10(3), 393-411.

Livingstone, S., & Helsper, E. (2007). Gradations in digital inclusion: children, young
people, and the digital divide. New Media & Society, 9(4), 671-696.

Manago, A. M., Graham, M. B., Greenfield, P. M., & Salimkhan, G. (2008). Self-
presentation and gender on myspace. Journal of Applied Developmental
Psychology, 29, 446-458.

Martin, A. J., & Dowson, M. (2009). Interpersonal relationships, motivation,

engagement, and achievement: Yields for theory, current issues, and educational
practice. Review of Educational Research, 79(1), 327-365.

Mayer, R. E. (2001). Multimedia learning. New York, NY: Cambridge University Press.

McLaughlin, M. W. (1990). The RAND change agent study revisited: Macro

perspectives and micro realities. Educational Researcher, 19(9): 11-16.

Nass, C., & Mason, L. (1990). On the study of technology and task: A variable-based
approach. In J. Fulk & C. W. Steinfeld (Eds.), Organization and Communication
Technology (pp. 46-67). Newbury Park, CA: Sage.

National Center for Education Statistics. (2006). Internet access in U.S. public schools
and classrooms: 1994-2005. Retrieved on December 4, 2008 from
http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2007020.
 134

Norris, P. (2001). Digital divide: Civic engagement, information poverty, and the internet
worldwide. New York, NY: Cambridge University Press.

O’Reilly, T. (2007). What is web 2.0: Design patterns and business models for the next
generation of software. Communications & Strategies, 65, 17-37.

Pasek, J., more, e., & Hargittai, E. (2009a). Facebook and academic performance:
Reconciling a media sensation with data. First Monday, 14(5),
http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2498/218.

Pasek, J., more, e., & Hargittai, E. (2009b). Some clarifications on the Facebook-GPA
study and Karpinski’s response. First Monday, 14(5),
http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2504/218.

Peter, J., Valkenburg, P. M., & Schouten, A. P. (2006). Characteristics and motives of
adolescents talking with strangers on the internet. Cyberpsychology & Behavior,
9(5), 526-530.

Pew Internet & American Life Project. (n.d.). Data sets: November 2007 – teens and
writing. Retrieved on December 4, 2008 from
http://www.pewinternet.org/PPF/r/69/dataset_display.asp.

Phillips, R. (2010, February 16). Facebook gripes protected by free speech, ruling says.
CNN. Retrieved from
http://www.cnn.com/2010/TECH/ptech/02/16/facebook.speech.ruling/index.html.

Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual
Review of Sociology, 24, 1-24.

Prensky, M. (2001). Digital natives, digital immigrants. On the Horizon, 9(5), 1-6.
Retrieved on January 13, 2009 from
http://www.marcprensky.com/writing/Prensky%20-
%20Digital%20Natives,%20Digital%20Immigrants%20-%20Part1.pdf.

Putnam, R. D. (2000). Bowling alone: The collapse and revival of american community.
New York: Simon & Schuster.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and
data analysis methods, 2nd Edition. Thousand Oaks, CA: Sage Publications.

Resnick, M. (2004). New styles of thinking for the era of organic networks. BT
Technology Journal, 22(4), 113-120.
 135

Rideout, V. J., Foehr, U. G., & Roberts, D. F. (2010). Generation M2: Media in the lives
of 8-18 year olds. Menlo Park, CA: The Henry J. Kaiser Family Foundation.

Roberts, D. F., & Foehr, U. G. (2008). Trends in media use. The Future of Children,
18(1), 11-37.

Royston, P. (2004). Multiple imputation of missing values. Stata Journal, 4(3), 227-241.

Ryan, A. M. (2000). Peer groups as a context for the socialization of adolescents’

motivation, engagement, and achievement in school. Educational Psychologist,
35(2), 101-111.

Salomon, G. (1984). Television is “easy” and print is “tough”: The differential

investment of mental effort in learning as a function of perceptions and
attributions. Journal of Educational Psychology, 76(4), 647-658.

Sandberg, S. How many friends can you have? Retrieved on May 29, 2009 from
http://blog.facebook.com/blog.php?post=72975227130.

Sharples, M., Graber, R., Harrison, C., & Logan, K. (2009). E-safety and web 2.0 for
children aged 11-16. Journal of Computer Assisted Learning, 25, 70-84.

Schmidt, M. E., & Vandewater, E. A. (2008). Media and attention, cognition, and school
achievement. The Future of Children, 18(1), 63-85.

Schmitt, K. L., Dayanim, S., & Matthias, S. (2008). Personal homepage construction as
an expression of social development. Developmental Psychology, 44(2), 496-506.

Schneider, B., Carnoy, M., Kilpatrick, J., Schmidt, W. H., & Shavelson, R. J. (2007).
Estimating causal effects using experimental and observational designs.
Washington DC: American Educational Research Association.

Schochet, P. Z. (2005). Statistical power for random assignment evaluations of education

programs. Princeton, NJ: Mathematica Policy Research Inc.

Schochet, P. Z. (2008). Statistical power for random assignment evaluations of education

programs. Journal of Educational and Behavioral Statistics, 33(1), 62-87.

Seiter, E. (2008). Practicing at home: Computers, pianos, and cultural capital. In T.

McPherson (Ed.), Digital Youth, Innovation, and the Unexpected (pp. 27-52). The
John D. and Catherine T. MacArthur Foundation Series on Digital Media and
Learning. Cambridge, MA: The MIT Press.
 136

Slavin, R. E. (2002). Evidence-based education policies: Transforming educational

practice and research. Educational Researcher, 31(7), 15-21.

Slavin, R. E. (2007). Educational research in an age of accountability. New York:

Pearson.

Stanton-Salazar, R. D. (1997). A social capital framework for understanding the

socialization of racial minority children and youths. Harvard Educational Review,
67(1), 1-39.

Stanton-Salazar, R. D., & Spina, S. U. (2000). The network orientations of highly

resilient urban minority youth: A network-analytic account of minority
socialization and its educational implications. The Urban Review, 32(3), 227-261.

Steinfield, C., Ellison, N. B., & Lampe, C. (2008). Social capital, self-esteem, and use of
online social network sites: A longitudinal analysis. Journal of Applied
Developmental Psychology, 29, 434-445.

Subrahmanyam, K., & Greenfield, P. (2008). Online communication and adolescent

relationships. The Future of Children, 18(1), 119-146.

Tien, F. F., & Fu, T. T. (2008). The correlates of the digital divide and their impact on
college student learning. Computers & Education, 50(1), 421-436.

Tufecki, Z. (2008). Can you see me now? Audience and disclosure regulation in online
social network sites. Bulletin of Science, Technology, & Society, 28(1), 20-36.

U.S. Department of Commerce. (2000). Falling through the net: Toward digital
inclusion. Retrieved on December 1, 2006 from
http://www.ntia.doc.gov/ntiahome/fttn00/contents00.html

Valkenburg, P. M., & Peter, J. (2009a). Social consequences of the internet for
adolescents. Current Directions in Psychological Science, 18(1), 1-5.

Valkenburg, P. M., & Peter, J. (2009b). The effects of instant messaging on the quality of
adolescents’ existing friendships: A longitudinal study. Journal of
Communication, 59, 79-97.

Valkenburg, P. M., Peter, J., & Schouten, A. P. (2006). Friend networking sites and their
relationship to adolescents’ well-being and social self-esteem. Cyberpsychology
& Behavior, 9(5), 584-590.
 137

van den Eijnden, R. J. J. M., Meerkerk, G. J., Vermulst, A. A., Spijkerman, R., & Engels,
R. C. M. E. (2008). Online communication, compulsive internet use, and
psychological well-being among adolescents: A longitudinal study.
Developmental Psychology, 44(3), 655-665.

Volman, M., & van Eck, E. (2001). Gender equity and information technology in
education: The second decade. Review of Educational Research, 71(4), 613-634.

Walther, J. B., Van Der Heide, B., Kim, S. Y., Westerman, D., & Tong, S. T. (2008). The
role of friends’ appearance and behavior on evaluations of individuals on
facebook: Are we known by the company we keep? Human Communication
Research, 34, 28-49.

Walther, J. B., Van Der Heide, B., Hamel, L., & Shulman, H. (2009). Self-generated
versus other-generated statements and impressions in computer-mediated
communication: A test of warranting theory using facebook. Communication
Research, 36(2), 229-253.

Wellman, B., Salaff, J., Dimitrova, D., Garton, L., Gulia M., et al. (1996). Computer
networks as social networks: Collaborative work, telework, and virtual
community. Annual Review of Sociology, 22, 213-238.

Whitehouse. (n.d.). President Obama launches “educate to innovate” campaign for

excellence in science, technology, engineering, and math (STEM) education.
Retrieved on February 16, 2010 from http://www.whitehouse.gov/the-press-
office/president-obama-launches-educate-innovate-campaign-excellence-science-
technology-en.

Williams, D. (2006). On and off the ‘net: Scales for social capital in an online era.
Journal of Computer-Mediated Communication, 11(2), article 11.
http://jcmc.indiana.edu/vol11/issue2/williams.html.

Wolak, J., Finkelhor, D., Mitchell, K. J., & Ybarra, M. L. (2008). Online “predators” and
their victims: Myths, realities, and implications for prevention and treatment.
American Psychologist, 63(2), 111-128.

Yan, Z. (2009). Differences in high school and college students’ basic knowledge and
perceived education of internet safety: Do high school students really benefit from
the Children’s Internet Protection Act? Journal of Applied Developmental
Psychology, 30, 209-217.

Ybarra, M. L., & Mitchell, K. J. (2008). How risky are social networking sites? A
comparison of places online where youth sexual solicitation and harassment
occurs. Pediatrics, 121(2), 350-357.
 138

Zhao, Y., & Lei, J. (2009). New technology. In G. Sykes, B. Schneider, & D. N. Plank
(Eds.), Handbook of Education Policy Research (pp. 671-693). New York:
Routledge.

Zywica, J., & Danowski, J. (2008). The faces of facebookers: Investigating social
enhancement and social compensation hypotheses; Predicting facebook and
offline popularity from sociability and self-esteem, and mapping the meanings of
popularity with semantic networks. Journal of Computer-Mediated
Communication, 14(1), 1-34.
 139

Appendix A

1. What is your District/Student ID Number: __________________________________

2. What is your teacher and class period: _____________________________________

3. Do you use MySpace? (Yes/No) __________________________________________

4. Do you use Facebook? (Yes/No) _________________________________________

5. We’d like to know a little about your school experience. Please indicate how true
each statement below describes you (check one on each line).

Not True At Not Very Sort of Very

All True True True
The most important thing for me
right now is improving my grade
point average.

I ask my teachers to clarify things I

don't understand well.

When I can't understand something

in my classes, I ask another student
for help.

Getting a good grade in my classes

is the most satisfying thing for me
right now.

I often come to class unprepared.

The most satisfying thing for me is

to understand what I learn in
school.

In my classes, I prefer material that

sparks my curiosity, even if it is
difficult to learn.
 140

Even if I have trouble

understanding something in class, I
try to do the work on my own,
without the help of anyone.

When studying for an exam, I

usually set aside time to study with
a group of classmates or my
friends.

When studying for my classes, I

often try to explain the material to
a classmate or friend.

I prefer classes that really

challenge me so I can learn new
things.

I work very hard in school.

I want to do well in my classes

because it is important to show my
ability to my family, friends, or
others.

I try to work with other students to

complete my school assignments.

6. On a scale from 1 to 5, how do you feel about your high school? (check one on each
line)

1 2 3 4 5
Important (1) or Not Important (5)

Involved (1) or Uninvolved (5)

Useful (1) or Not Useful (5)

Enthusiastic (1) or Not Enthusiastic

(5)

Stimulated (1) or Not Stimulated (5)

 141

Looking forward to it (1) or Dreading

it (5)

Invigorated (1) or Not Invigorated (5)

Excited (1) or Not Excited (5)

Helpful (1) or Not Helpful (5)

Interested (1) or Uninterested (5)

7. How many teams, clubs, school groups, or extracurricular activities are you a part of?
(check one)

One Two Three Four More than Four

8. How strongly do you agree with these statements?

Strongly Agree Disagree Strongly

Agree Disagree
A. There are people in my high school I
trust to help solve my problems.

B. There is someone in my high school I

can turn to for advice about making very
important decisions.

C. There is no one in my school that I feel

comfortable talking to about personal
problems.

D. When I feel lonely, there are several

people in my school I can talk to.

E. If I needed an emergency loan of $500,

I know someone in school who I can turn
to.

F. Interacting with people online makes

me interested in things that happen outside
 142

of my town.

G. Interacting with people online makes

me want to try new things.

H. Talking with people online makes me

curious about other places in the world.

I. Talking with people online makes me

feel like part of a larger community.

J. Interacting with people online makes me

feel connected to the bigger picture.

* Only answer the following questions if you were in a class period that used the social
networking site.

9. Did you use the site often or rarely? _______________________________________

10. Did you find the social networking site useful in your classroom?

11. What do you think could be done better next time to motivate students like yourself to
use a social networking site in your classes? (answer in the space below)

Do you think social networking sites should be used in school? What kinds of school

activities would you use a networking site for? (answer in the space below)