Download as pdf or txt
Download as pdf or txt
You are on page 1of 30

Determinants of Helping Behaviors in Online Groups: A Conceptual Model

Mani R. Subramani
Associate Professor
msubramani@csom.umn.edu

Naren Peddibhotla
Doctoral Program
npeddibhotla@csom.umn.edu

3-365, Information and Decision Sciences Department


Carlson School of Management,
University of Minnesota
321, 19th Ave South
Minneapolis, MN 55455.

Last revised: Aug 5, 2004

Accepted for presentation at


Academy of Management Conference, OCIS Division
August 6-11, 2004, New Orleans, LA.
Determinants of Helping Behaviors in Online Groups: A Conceptual Model

Abstract

There are at currently least 8 million unique users of over 50,000 online groups on the Internet.

These are forums where individuals discuss and share information about a wide spectrum of work-

related and non-work topics using email or the web. These forums have a high level of message

traffic - over 150 million messages were posted in Usenet groups in 2000 (Smith 2002). The

occurrence of helping behavior observed in these online groups raises several questions. What are

the characteristics of helping - what kind of help do people provide online? What are factors

influencing helping? What are the mechanisms supporting such prosocial behavior and how do

features of the context of online groups enable or inhibit helping behavior?

This paper draws on prior work on online groups, literature on helping in interpersonal situations,

and research on the dynamics of groups to propose a model explaining the helping behavior of

participants. It highlights that the size of the group and the diversity of the group have complex

effects on the nature of helping behavior. It suggests that the creation of ancillary resources (such as

FAQs and repositories of content supplementing the site) influences the nature of helping behaviors

information exchange is inhibited while knowledge creation and sharing is enhanced. This paper

also highlights the unrecognized value of occasional participants and passive participants in

influencing behavior and the complementarity in the roles of central and peripheral participants.

August 5 2004 2
Determinants of Helping Behaviors in Online Groups: A Conceptual Model

1. Introduction

There are over 50,000 online newsgroups on the Internet (Smith 2002). The intents of these

groups, typically signaled by the name selected for the group encompass a wide variety of topics.

Examples are groups discussing and sharing information on professional issues like computer

programming, physics and anthropology and groups discussing personal interests and hobbies like

quilting, Indian classical music, politics and hang gliding. These groups are forums completely open

to all comers on the Internet. Most groups have few, if any, membership criteria, and anyone can

join a group and begin participating. In these online groups, participants are invisible till they make

their presence known by posting to the group. The only way participants learn about others on the

group is through their actions on the newsgroup such as posting questions, responding to questions,

maintaining the repository of frequently asked questions (FAQ) and expressing their views through

postings on the newsgroup. Participants usually become familiar with other members over time,

even if they only know each other by their screen names. Many groups also evolve norms of

courtesy and conduct and regular members of these groups come to identify closely with them

(Sproull and Faraj 1995). The maintenance of these public forums also involves social management

e.g. to discourage inappropriate use, to enforce norms of courtesy and groups typically evolve

mechanisms to accomplish these tasks on an ongoing basis. Online groups are also referred to as

online communities or online social networks on account of these social characteristics (Butler,

Sproull, Kiesler and Kraut 2002, Faraj and Sproull 1995).

These online groups have lively and often spirited discussions on issues of interest to

members. They are also forums where individuals ask for advice or comments and request help in

solving problems with the expectation that other group members would willingly make the effort

and take the time to provide help and share their expertise. The technology used to implement

Online groups have features to support simultaneous discussions on multiple topics. Messages
August 5 2004 3
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
posted to the group are organized by topic (signaled by the header used for the message) and

subsequent responses are threaded hierarchically. The unique feature of online groups is the high

level of public visibility of actions on these forums to the large, mainly invisible audience. All action:

requests for help, postings providing help, and views expressed on the forum are visible to the entire

set of participants. Posts to forums often archived periodically, organized by topic, for later

reference. Prior studies suggest that online groups can become very important components of the

social and professional lives of participants (Sproull and Faraj 1995). Participants report a variety of

motivations for joining and frequenting these online groups: moral and emotional support, advice

and information on various issues, and affiliation and connecting to others with similar tastes and

interests (Finholt and Sproull 1990, Sproull and Faraj 1995, Wasko and Faraj 2000). Clearly, online

groups are complex social collectives and the observation of actions in online groups raises

questions of motivation (why, in the absence of direct payoffs from helping do people help?), of the

characteristics of helping (what kind of help do people provide online?), and of mechanisms

supporting it (how do features of the context of online groups enable or inhibit it?).

The action of responding to requests for help and engaging in group maintenance tasks is

prosocial - intended largely to benefit people other than the respondent. A broad range of prior

research in social psychology has examined prosocial behavior in contexts where both the helper

and the help-seeker interact face to face (Batson 1998). However, the results of prior work on

helping in conventional contexts are limited in their applicability to helping behavior online. Some

of the factors most salient in influencing prosocial behavior in face to face contexts such as the

physical characteristics of help seeker, the physical setting of the request for help and the physical

proximity of the help seeker (Betancourt 1990, Brown and Smart 1991), are absent or are

significantly muted. For instance factors such as cultural similarity and communal bonds that are

August 5 2004 4
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
known to enhance helping are difficult to ascertain online. It is also easier to ignore a request for

help online than in a face-to-face context where the help seeker is physically present.

Moreover, online contexts have unique characteristics not shared by face-to-face contexts of

prosocial behavior: narrow focus, asynchronous interaction and searchable records of activity. The

intent of online groups is usually narrow and focused, concerning a particular kind of activity e.g.,

programming Oracle databases, K-12 education or a narrowly specified interest e.g. flying aerobatic

aircraft. All discussion, and the seeking and receiving of help occur asynchronously whenever

participants log on and post. These can often stretch over several days, weeks and even months.

Ongoing interactions can occur between individuals who know little of each other beyond the fact

that they participate in the group. Online groups often evolve mechanisms to create a searchable

record of past interactions though the compilation of FAQs. Further, online groups require

ongoing maintenance activities for their sustenance e.g. the discouraging of inappropriate use,

recognizing outstanding contributors, stepping in to de-escalate disputes between individual

members etc. These administrative duties are generally performed voluntarily by members of the

group.

The literature examining helping behavior in online groups is relatively sparse. The study of online

helping in Internet groups by Wasko and Faraj (2000) suggests that altruism, generalized reciprocity

and community interest created by ongoing interaction of the members of these online groups are

important motivations for participation. Butler et. al (2002) find that a mixture of altruism, social

benefits and the desire for visibility were factors motivating contributions to the maintenance of

online communities. Constant, Kiesler and Sproull (1996) examined a more restricted context - an

August 5 2004 5
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
organization network1. Their examination of the use of email for help seeking suggests that

citizenship behavior (the desire to benefit the organization) motivated helping.

However, while each of these studies examines specific instances of helpful behavior in

online groups, a comprehensive picture of factors determining prosocial behavior in online groups

has yet to emerge. A broader understanding of these factors is an important issue with a variety of

consequences for public networks online. Online groups have features similar to those of public

goods (Kollock and Smith 1996). It is important to understand the set of issues that combine to

sustain them as dynamic forums for discussion, learning and knowledge exchange so that such

forums do not inadvertently deteriorate and the value created by them lost as a result. Further,

understanding helping in online groups can help inform firm efforts to encourage helping and

knowledge sharing by employees in proprietary networks within organizations.

In this paper, we apply the findings of research on online groups and research on helping

behavior in interpersonal situations to the context of online groups. In doing so this paper

contributes to the literature in four ways. First, it highlights the complex relationship of size and

diversity on interactions in online groups. Second, it links the unique features of online groups, such

as the ability to create ancillary resources to supplement group interaction and the nature of group

membership, to the nature of helping on these networks. Third, it discusses the distinctive

determinants of helping behavior and the usefulness of online groups to participants. Finally, it

emphasizes the importance of peripheral participation that is largely invisible in online groups and its

part in complementing visible forms of active participation.

1
Online groups are typically open membership groups, bringing together individuals on the basis of shared
interests. Organizational networks on the other hand are generally proprietary forums limited to employees of the
firm.

August 5 2004 6
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
2. Helping in Online Groups2:

Interaction in online groups comprises posts by individuals and subsequent responses to

these posts by others. Responses to posts reflect the willingness of members to help or to share

their opinion with the original poster. The nature of responses posted to messages thus provides a

means to assess the characteristics of helping behaviors in online groups3.

Helping in online groups is conventionally viewed as comprising information sharing in

response to requests for assistance e.g. by responding to questions such as does anybody

know.(Constant, Kiesler and Sproull 1996). Researchers also recognize that the sharing of

perspectives with participants in the group and the discussions and debates occurring in these

forums on different issues reflect the creation of new knowledge and knowledge sharing (Wasko and

Faraj 2000). This is consistent with the suggestions of Nonaka (1990) that discussion and discourse

between members of groups are central mechanisms enabling the dissemination of knowledge, the

elaboration of existing knowledge and the creation of new knowledge. We thus focus on the two

dimensions that reflect the nature of helping in online groups: a) information exchange and b)

knowledge creation and exchange. This parallels the broad distinction between providing a response

that is an answer and providing a response that is more than answer noted in contexts of interpersonal

helping (Cross 2000).

Information Contribution: We characterize the help provided in response to posts requesting

assistance, e.g. I need the postal code of the Olympic Games Village in Barcelona as information

contribution. Individuals posting responses to such requests help by answering the question or

providing the information needed. Providing the information may also involve a certain level of
2
We limit our scope to prosocial behavior evident from members willingness to take the time and effort to respond
to posts in online groups. We do not examine community building activities in groups such as those discussed by
Butler et. al (2002) that also represent instances of such behavior.
3
We limit our consideration to help rendered by posting comments or responses to messages by other participants.
Group members may also help by contributing posts that are not responses e.g. informational announcements of
potential interest to members e.g. Firmware v1.4.1 for SMC7004AWBR now available for download at the SMC
website.

August 5 2004 7
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
judgment on the part of the helper. For instance, a user on comp.human-factors, a group devoted

to computer interfaces posted the following question:

What would be a good place to look for usability /human-computer interaction


problems/issues pertaining to hand-helds/PDA's ???
One of the responses to this was:

The proceedings of Mobile HCI 2001 are online:


http://www.cs.strath.ac.uk/~mdd/mobilehci01/procs/
In general, information sharing occurs when questions pertain to issues that have already been

considered and problems for which solutions exist. Questions arise (creating the need for posts)

when the information is either not well known or the questioner is not aware of it. Members of the

group possessing the information or those who can access the information respond, and in their

response, often point to sources for further information such as book chapters and articles.

Knowledge Contribution: We characterize responses to posts involving issues that have no

answers as reflecting Knowledge Creation and Exchange. These responses involve issues on which

there are multiple perspectives held by individuals; the validity and generality of different

perspectives can be determined only through discussion and debate. Consider the following message

on comp.human-factors:

Subject: Is it possible to cultivate trust relationship online?


Let us discuss online behaviour from a specific aspect-trust. In comparison
with the current network, the size of the early Internet was tiny, and most
users tended to be academics or to be involved in research or to have a
technical background. It is likely that they were familiar with each other,
at least with their names. They shared interests within user-groups and this
climate of trust eases cooperation among professionals. However, the mix of
Internet users is increasingly being diluted from its specialist origins by a
vast influx of additional users, who are from a huge range of different
backgrounds. Vulnerability of Internet user during online communication has
gone beyond what we expected before. Potential harm from strangers to us
normally results in loss of time, information, privacy and property. Under
this circumstance, how can we know if the trust we choose to give or withhold
is warranted?

This is clearly an issue with no ready answer and this post received five responses providing a variety

of perspectives on the issue. In such discussions, participants may choose to comment on one of

August 5 2004 8
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
the responses rather than respond to the original message. Clearly, the behavior reflected in

responses reflect knowledge sharing and lead to the dissemination of knowledge. As Nonaka (1994)

suggests, each response broadens or elaborates the understanding of the issue and helps create new

knowledge on the issue. Discussions on the forum benefit not just the active members participating

in the interaction but also the passive participants, observing action passively from the sidelines4

reading posts but not actively contributing to discussions.

As the contributions of individuals are visible to all participants, active contributors in online

communities derive visibility payoffs in the form of recognition as being informed, knowledgeable and

gain the respect of peers in the group (Butler et. al. 2002). The two forms of helping information

sharing and knowledge contribution, are associated with differing levels of visibility payoffs.

Contributions of information are likely to have lower payoffs than contributions of knowledge for

several reasons. To make contributions of information, individuals need only to look up a textbook,

manual or other authoritative source containing the details required a task any competent member

of professional groups should be able to perform5. On the other hand, contributions of knowledge

typically require judgment and the application of expertise. Knowledge contributions involve

contexts where clearly defined rules do not exist for selecting, evaluating, and combining inputs to

arrive at conclusions. Clearly, the capability to exercise judgment and expertise developed from

intimate knowledge of issues are both usually in shorter supply and less widely distributed than

information. Knowledge contributions are therefore likely to be associated with higher levels of

visibility payoffs than information contributions.

4
This group, which by some accounts comprises over 90 percent of the membership (Zhang and Storck 2001) are
viewed charitably as the silent majority (Butler et. al 2002) or in a more sinister vein as lurkers (Finholt and Sproull
1990).
5
The information can be looked up only if individuals know where to look. An awareness of sources of information
is a key attribute distinguishing trained novices from laymen .

August 5 2004 9
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
3. Factors Influencing Helping Behaviors

In the sections that follow, we draw on prior literature in highlight the factors influencing

the level of helping behavior reflected in the level of information exchange and the level of

knowledge sharing in online groups. The overall theoretical framework is illustrated in Figure 1.

Prior research on prosocial behavior, social psychology and group dynamics suggest a variety of

features that are likely to influence the nature of helping behavior in online contexts.

3.1 Group Size and Helping Behavior

While size has been suggested as a factor affecting interactions in collocated groups (Batson

1998), prior research on online groups adopts the view that one of the features of online communities is that

its population size does not matter in the conduct of many activities (Butler et. al 2002, page 11). In contrast, we

suggest that the relationship of size to outcomes is complex and that group size (in terms of the

number of members) is an important factor influencing the nature of helping in online groups.

Drawing from prior literature, we highlight the complexity of this link and highlight different strands

of reasoning suggesting a positive association between group size and helping : expanding resource

access, greater social learning and greater peripheral participation. We also highlight the mechanisms

suggesting a negative association of group size and helping: greater diffusion of responsibility, information

overload and wearying of active contributors.

Expanding resource access by hardwiring weak ties: Online groups are argued as strengthening and

hardwiring weak ties (Pickering and King 1992) by facilitating communication exchanges.

Online groups allow individuals to reach out to a larger group of potential helpers than practically

feasible using conventional means of 1-1 communication such as face-to-face interactions and

telephone calls. It requires no more effort to post to a group with several thousand members, than

that to send email to one acquaintance. Thus larger networks would afford individuals access to the

resources of a larger group than smaller networks. From this perspective, the size of the online

August 5 2004 10
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
group the number of people signed up as members of the group or the number of participants in a

group has positive effects on helping, an increase in the size of the group.

Social Learning: From this perspective, larger networks provide greater opportunities for participants

to observe prosocial behavior and a larger variety of positive group oriented behaviors that

newcomers can model (Bandura 1977). Networks with a large number of participants have a larger

number of messages eliciting responses than smaller networks, ceteris paribus. Larger networks also

have, by the same token, a larger pool of potential helpers who respond. Participants in large

networks are therefore likely to observe a larger volume and a greater variety of help seeking,

information exchange and knowledge sharing than in smaller networks. Therefore, larger social

networks, through the processes of social learning and socialization of members are likely to be

contexts with greater helping behavior and contexts that are generally more helpful than smaller

networks.

Legitimate Peripheral Participation: The extent to which environments provide mechanisms for new

entrants and individuals considering participation to observe work being performed and be within

earshot of discussions between experts as they resolve problems is considered central to learning and

competence development in a variety of contexts (Hutchins 1990). Online groups are forums where

individuals from a variety of backgrounds sharing common interests e.g. in Human Computer

Interfaces (e.g. comp.human-factors), the politics of a certain region (alt.politics.lebanon), a

philosophical belief (alt. atheism) or a hobby (such as rec.music.Indian-classical) can

congregate to seek information, share opinions, mobilize to address worthy causes and act in ways

that generally benefit the group. Online groups6 are uniquely inclusive, individuals sharing interests

with the group can participate irrespective of their backgrounds, affiliations or physical location.

6
It is possible to be restrictive in membership with a moderator screening individuals allowed to join a Listserv
group. However, a majority of groups online appear to be open and to the best of our knowledge, no statistics are
available on the number of closed and open groups.

August 5 2004 11
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
Further, online groups are generally transparent forums with a culture of open-ness and consensual

decision-making. Unlike conventional groups where newcomers face social barriers (key decisions

are often made in informal meetings of influential members) and practical constraints preventing

their being privy to how the group functions, all activities of members in online groups are

completely open, visible to all members with publicly records available record of all activity for

review. Online forums thus provide a venue for novices and others wanting to learn about issues

discussed in the group to hang around and learn from observing the proceedings of the group and

participate when that they feel comfortable doing so. The period of observation from the sidelines

can be an important part of participation because many groups evolve specialized vocabularies and

nuances related to even standard technical vocabularies. For instance, consider this exchange in an

UnixNT-L listserv that requires not just technical knowledge but also knowledge of the specifics of

NT releases (nt40, sp3), local vocabulary (you will probably have to telnet, scripted through expect or

something else) and cryptic references to firms or groups whose identity is clarified cryptically ms

etc. :

>Does anyone know how to get rsh to work between unix and NT servers? I have been racking my brains
>on this one from quite some time and can't get anything to work.

sure. nt comes with rsh [but not rlogin]. type rsh <unixhost> -l root echo
hello and see if it works. if your /etc/hosts and .rhosts file are like they should be, it will work fine.

if you want to rsh from unix to nt, that is different. you will probably have to telnet, scripted through expect or
something else. we [ms] just released services for unix, which includes a telnet server for nt40 sp3 and
above, so nt/unix interop should be decent.

Lurking, viewed as time spent in the background to become familiar with the group is therefore

legitimate peripheral participation (Wenger 1988) that serves to help novices develop competence and

gain confidence and allows them to come abreast of developments and current opinions in the

domain of the group at their own pace. In general, larger networks with greater network message

traffic that afford a wider horizon of observation the variety of task performance that is observable to

August 5 2004 12
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
novices (Hutchins 1990) are therefore more useful in this regard than networks that are smaller in

size.

The development of competence by novices also has the positive effect of broadening the

distribution of expertise in the group. Questions or issues requesting comments posted to larger

forums are therefore more likely to receive a larger number of responses as well as more useful

responses because of the there are likely to be more participants with the capability to help.

Overall, these arguments provide the basis to suggest that larger networks are likely to be

contexts marked by higher levels of helping behavior and contexts that are likely to be more useful

for participants than smaller networks.

P1: The level of helping behavior in the group is positively related to the size of the group.

However, it is also likely that network size may be detrimental to helping and that larger networks

may therefore have less helping behavior than those smaller in size.

Diffusion of Responsibility: While the number of potential helpers increases with the size of the

network, the level of motivation of individuals to help is an important factor determining the

assistance that is actually provided. Prior research suggests that in the presence of other potential

helpers, the level of individual motivation to help diminishes; a phenomenon termed the bystander

effect (Latane and Darley 1970). The stabbing death of Kitty Genovese is often used to illustrate this

phenomenon (Latane and Darley 1970). Kitty, a young woman, was attacked on her way back from

work in the alleyway of the apartment complex where she lived in New York City. Her screams for

help as she was attacked multiple times over a 45 minute period were registered by at least 38

apartment residents. However, not one of them performed the simple helping act of picking up the

phone to call the police (Latane and Darley 1970).

These arguments highlight that helping behavior is related to an individuals feeling of

responsibility for providing help. An increase in the number of other potential helpers in larger
August 5 2004 13
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
networks leads to the diffusion of the responsibility that each individual feels to help. Overall, this

can cause the level of helping behavior in larger networks to be lower than the level in smaller

networks.

Information Overload: An increase in the size of the online group, by increasing the information load,

thus can reduce the level of helping by members. Online groups, in general, have a large number of

messages posted on them daily. For instance, in March 2002, the newsgroup comp.sys.ibm.pc.games.rpg

had an average of 180 messages posted per day. The high message volume on a daily basis on high

traffic online groups imposes a considerable workload and information-processing burden on

participants (Konstant et. al 1997). Information overload adversely affects the ability of individuals

to help in several ways. When presented with a large number of parallel conversations on the site,

participants may fail to notice postings relevant to their interests and expertise that they may be

willing to respond to. In essence, these messages get lost amid the large volume of messages of no

interest to the individual - a phenomenon termed the data clutter problem (Woods, Patterson, Roth,

Christoffersen 2000). In addition, participants may reduce their involvement with the online group

in an attempt to cope with the information overload, leading to a reduction in their level of helping

in the forum. Overall, the effect of information overload is analogous to the diminishing of

involvement and the reduction in the helpfulness of individuals in large cities suggested by the urban

overload hypothesis (Steblay 1987).

Wearying of active contributors: Online groups generally have a small core of enthusiastic participants

who are the most active contributors on the forum. The bulk of the membership of online groups

comprises passive participants who observe the messages posted on the group- such participants are

often described as lurkers. In a study of posts over one month to a travel forum with 1065 users,

Zhu and Storck (2001) found that 78 active users provided 4777 responses. In effect, the bulk of

August 5 2004 14
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
the membership of this group relied on this small fraction (about 7 percent of the total) to provide

the information and knowledge shared in the group.

As the group size increases, it is probable that the number of experts willing to provide

assistance does not increase proportionally. This can thus lead to the core group of active

responders being overloaded. The considerable demands on the time and attention can be

burdensome and extract a significant toll on active contributors (Butler et al 2002). These

individuals may just become weary of continuing to cope with increasing demands on their expertise

and resources and reduce their level of helping as group size increases. This therefore suggests a

negative relationship between the size of a group and the extent of helping in online groups.

Overall, these arguments suggest that larger groups are likely to be characterized by lower

levels of helping behavior than smaller groups.

P1b: The level of helping behavior in the group is inversely related to the size of the group.
3.2 Membership Diversity

In general, the diversity of members contributes significantly to group outcomes (Williams and

OReilly 1998, Pelled, Eisenhart and Xin 1999). The diversity of membership of groups is vital to

allow individuals to encounter a variety of viewpoints other than their own and for members to

learn, develop and refine their own perspectives through interaction with others with different

views. This is highlighted by Franke and Shah (2001) who studied four sports related communities

sailplaning, canyoning, sailboarding and handicapped cycling. A mix of members with varied

interests enthusiastic sportsmen (who try out equipment and provide ideas for new products),

innovators (who think of new ways to enhance performance in the sport), manufacturers (who take

these ideas and turn them into products) and sportsmen (who test products and provide feedback) is

an important feature enabling the groups to generate and diffuse innovations relating to sports

equipment.

August 5 2004 15
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
The relationship between the diversity of an online groups membership and the level of

helping is likely to be complex. If an online group is defined narrowly (e.g. as being for users of high

performance snowboards), it is likely that the members of the group are homogeneous. With a high

level of similarity in interests, backgrounds and the level of expertise of members, there is little that

another member is likely to know or be aware of that an individual does not already know -

opportunities for information sharing and knowledge sharing are therefore both likely to be limited.

This would be reflected in a low level of helping behaviors in the group. On the other hand, an

extremely diverse group is likely to also find it difficult to interact and exchange ideas using the lean

medium of text based messaging (Daft and Lengel 1986).

To the best of our knowledge, the role of diversity in online groups has received little

attention by researchers. Drawing form the literature on team diversity, we suggest that two aspects

of diversity are likely to be important in influencing the level of helping in online groups: the diversity

of expertise in the group (Rau 2000, Libby, Trotman and Zimmer 1987) and the cognitive diversity of

group members (Miller, Burke and Glick 1998).

Diversity of Expertise: A group with a membership that is diverse in terms of the level of expertise in

the particular domain is likely to be most strongly associated with helping. Consider a group

focused on C# programming consisting primarily of novices. Questions raised by group members

requiring expertise are unlikely to be answered or likely to be answered simplistically by other peers.

Clearly, such a group is unlikely to exhibit much helping behavior and not be generally considered

useful by participants. Similarly, an online group consisting primarily of highly proficient experts is

unlikely to be the context for much helping either as these individuals may have few doubts worth

clarifying in the group. A group with a diverse mix of expertise e.g. comprising novices,

programmers with an intermediate level of expertise and experts with considerable experience and

expertise is likely to be the characterized by high levels of helping. In such a group, the novices can

August 5 2004 16
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
be the source of simple informational questions and in the course of asking questions - may naively

frame familiar problems in novel ways and spark exciting discussions. Intermediate level

programmers who would develop competence by tackling unfamiliar problems would be active in

providing responses to such posts. Experts in the group with superior conceptual and problem

solving skills could potentially contribute responses to responses - commenting on more efficient ways to

solve problems than those suggested by other helpers, suggest new approaches to problems etc.

Expertise diversity can be conceived in terms of the diversity of the domains of expertise as

well e.g. expertise in designing products and expertise in using the products (Franke and Shah 2001).

In general, online groups with greater levels of expertise diversity are likely to exhibit higher levels of

helping behavior than those with lower levels of expertise diversity.

Cognitive Diversity: The ability for individuals to effectively communicate requires the sharing of

vocabulary and meaning (Rau 2000) and in the absence of a shared cognitive context, the level of

helping and the usefulness of the group are likely to both be low. The transfer of information and

knowledge is facilitated when members share a common vocabulary (Nonaka 1990). Some

components of knowledge are described as being sticky as they can be exchanged only among

individuals sharing similar vocabularies and cognitive structures (Szulanski 1996). This suggests that

a high level of diversity is likely to impede information and knowledge sharing while low cognitive

diversity is likely to be associated with freer exchange of information and more productive discourse.

Cognitive diversity is particularly detrimental to online groups as exchanges between

members are mediated by a low richness medium (asynchronous text messages) that provides little

support for bridging differences in meaning and interpretation. Overall, expertise diversity has

positive effects on helping and overall helpfulness while cognitive diversity has negative

consequences. These arguments suggest the following propositions:

August 5 2004 17
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
P2a: Online groups with higher levels of expertise diversity are characterized by greater levels of helping than groups
with lower levels of expertise diversity.
P2b: Online groups with lower levels of cognitive diversity are characterized by greater levels of helping than groups with
higher levels of cognitive diversity.
3.3 Ancillary Resources

Most online groups maintain repositories of ancillary information to supplement the information

and knowledge sharing online. The FAQ7 is the most common instance. FAQs are created by

compiling (or anticipating) answers to questions frequently raised by novices in the group and are

repositories of basic domain specific information in Q&A format. FAQs are generally updated and

maintained by volunteers. The FAQs is esseantially a device that makes routine information

available to those needing it without them having to post a question to the forum. It is generally the

norm that participants look up the FAQ to verify if a question has been archived in the FAQ before

posting it to the group. Members are encouraged to minimize the use of the online forum for

information that can be obtained by looking up the FAQ. The intent underlying the creation of

FAQs is to conserve network bandwidth and to not impose on members time and attention for

basic questions.

Online groups may also create elaborate ancillary resources that are broader in scope than

the FAQ. For instance, SEMNET a LISTSERV group for the discussion of Structural Equation

Modeling issues, maintains a freely accessible group homepage at www.gsu.edu/~mkteer/semnet.html.

This site contains a rich set of resources such as lists of books on SEM, comparison of different

commercially available programs for structural analysis and details of papers on structural equation

modeling methods. The availability of these resources is periodically publicized on the online group.

7
The creation of FAQs is a tradition in Usenet groups. Over the years, custodians of Usenet FAQs have informally
evolved elaborate policies regarding the creation and maintenance of these documents. It is an informal policy for all
groups to archive their FAQs at a central site rtfm.mit.edu to make these documents freely available.

August 5 2004 18
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
These ancillary resources are generally referenced in responses posted on online groups and users

pointed to material available in these ancillary resources.

The availability of ancillary resources changes the nature of requests in online groups and

consequently influences the nature of helping behavior in several ways. First, as looking up ancillary

resources prior to asking questions becomes a group norm, the incidence of requests requiring

information sharing is reduced. Further, nave posters posting questions answered in the FAQ often

receive terse, publicly posted responses, suggesting that they should do their homework before

taxing the resources of the group. The availability of ancillary resources thus reduces the legitimacy of

asking informational questions. This inhibits novices those most likely to face the questions from

posting them to the group and reduces the incidence of information exchange. Second, payoffs

from responding to informational requests are diminished as they are less valued as robust

searchable FAQs and resources are available. Participants are thus likely to focus on making

knowledge contributions and the incidence of this type of helping is likely to increase.

Further, with the reduction in the clutter from the repeated requesting of broadly known or

accessible information, requests for knowledge contributions become more salient increasing the

level of such helping. Overall, the availability of ancillary resources tends to reduce the level of

information contribution while enhancing the level of knowledge contribution. This leads to the

following:

P3a: The degree of elaborateness of a groups FAQ is inversely related to the level of information contributions.
P3b: The degree of elaborateness of a groups FAQ is positively related to the level of knowledge contributions.

3.3 Peripheral Members and the Nature of Interactions

The bulk of the membership of online group comprises members who play a passive role they

post questions or comments intermittently on the group but do not generally participate in the

group. This silent majority is sometimes referred to as being drive-by posters as they appear, seemingly
August 5 2004 19
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
out of nowhere to contribute a post and recede into the background to not be seen again for a while.

In general, this set of participants on the sidelines is seen as free riding on the efforts of active

participants and even described in sinister terms as lurkers (Finholt and Sproull 1990). There has

been very little systematic examination of the role of the silent majority and their influence on the

nature of interactions.

A more balanced view of this group emerges if they are recognized as peripheral participants

(Wenger 1998). We suggest that this set of silent observers represents the talent pool from which

active participants in the group can emerge as individuals become socialized, develop confidence and

gain confidence from legitimate participation on the periphery. The one study that examined

peripheral members on a group pertaining to travel (Zhang and Storck 2001) found that they

generated a substantial number of questions. It is hardly surprising that these participants are more

likely to contribute initial posts in the form of questions rather than to provide responses. These

peripheral members, much like novices and apprentices, are likely to incrementally work their way

towards more central participation in the community. It is likely that they begin by posting

questions and making information contributions. Novices are perfectly capable and in fact more

suited than experts to contribute information as they often possess more information and are more

abreast with developments in the field than experts (Shanteau 1992). Further, it is rational for

experts to prefer to engage in knowledge exchange than information sharing as this represents a

more effective use of their time and talent. In contrast, motivated novices on the periphery of the

group may welcome information exchange as a route to become more central in the community. In

essence, this represents a natural division of labor in online communities. It is probable that

peripheral members performing the bulk of the responding to informational requests in many

groups.

August 5 2004 20
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
The overall contributions of peripheral members is likely to be under-appreciated as there

are a much larger number of peripheral members in the group (than central members) each of

whom is likely to restrict their contribution to a small set of issues they feel competent in. We

believe that a systematic examination of the contributions by individuals in online groups over time

is likely to throw more light on this issue. These arguments lead to the proposition that:

P4a: The proportion of information sharing message associated with peripheral members is larger than the proportion
attributed to central members of online groups.
P4b: The proportion of knowledge sharing messages associated with peripheral members is smaller than the proportion
attributed to central members of online groups.
3.4 Social Capital Effects and Helping8

The level of helping by members in groups is influenced by a range of mechanisms that are broadly

classified as reflecting the level of social capital. For instance, group members expect that their

efforts to help others will be remembered so that their prosocial behavior will be reciprocated.

Members expect that irresponsible behavior in the group will be inhibited by a cumulative memory

of incidents (posting off-topic messages, using inflammatory language in posts, attacking others

personally in messages etc). Overall, group members expect that positive, prosocial behavior in the

group would enhance their standing and credibility creating social capital that would facilitate their

ability to rely on others and seek help in future periods.

A pre-requisite for social capital effects to operate is the existence of mechanisms to

preserve the history of prior actions by individual participants. This function is normally

accomplished in contexts of stable membership by relying on the memory of individuals. However,

membership in online groups is inherently fluid and ill defined (Butler et al 2002) on account of the

nature of the underlying technologies9. The membership of Usenet groups is the set of people who

8
We focus on examining the factors enabling social capital effects to influence helping .
9
To join a listserv, prospective members are required to send an email message to a specified Listserv address. No
act by users is required to join a Usenet group, individuals need to just click or type the name of a group to examine

August 5 2004 21
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
read the content posted on them - membership is open to all users of the Internet who care to view

the content. Usenet groups are similar to bulletin boards in the physical world, the users are all

individuals who post notices as well as those who care to stop and read notices. There is no

mechanism available to identify users who merely read notices and never post and further, no way to

track viewership except by inferring it when an individual adds a posting to the board. Membership

in a Listserv is more tangible in the sense that the users email address is listed along with the group

of other members and posts are emailed to the entire group. Membership in Listservs can be

ascertained from the list maintained at the list server.

The overall nature of group membership, whether it is relatively stable or whether it is

constantly in flux with very little overlap in membership from period to period has a significant

influence on the level of shared history of group members and the ability for social capital effects to

operate in the online group. This is therefore a key factor determining the extent to which helping is

influenced by social capital.

Factors that keep individuals returning to read posts on a regular basis are likely to be the

key determinants of the stability of group membership. While the examination of Usenet posts

suggests that there are regulars who frequent online groups (Smith 2002), few prior studies have

directly investigated factors that keep users committed to spending their time and energies reading

and posting to the group.

We suggest two factors that are likely be influential in enabling social capital effects in online

groups: a) the extent to which the group has a unique identity as the location for information on a

particular topic and b) the extent to which the membership of online groups overlaps with

membership in groups in the physical world . For instance, the members of ISWORLD (a Listserv

group) overlaps significantly with the overall population of faculty and doctoral students in

the contents and post messages if they choose to do so. Being a part of a Usenet group is quite like joining the
listenership of a radio station, anyone who tunes in on the frequency is a listener.

August 5 2004 22
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
Information Systems faculties around the world, a relatively small community where individuals

know each other. The overlap enables social capital mechanisms to operate in the online group. We

discuss each of these factors in details below.

Uniqueness: The consumer choice literature (Brent and Gibbs 1993) and the literature in marketing on

brand loyalty (Chaudhury and Holbrooke 2001) both point to the uniqueness of a service as a key

determinant of both customer choice, the ability to retain customer loyalty and retention. We

suggest that the level of uniqueness of a group is important in the online setting as well. In view of

the large number of alternative forums available online and the ease with which users can access

content posted on any or all of them, the uniqueness of an online group is likely to enables it to be

the preferred choice by users for content related to that topic and therefore lead to a more stable

participant population. For instance consider online forums relating to two programming languages

PROLOG and LISP. comp.lang.prolog is one of four forums on the net for discussion of

issues related to Prolog while comp.lang.lisp is one of nine groups for discussions

of lisp. Social capital effects are likely to be stronger than in the prolog group than in the lisp

group, ceteris paribus, as it is likely that members have more options in the case of lisp.

Overlap between Online and Physical groups: The overlapping of online groups with physical groups is

likely to influence the retention of members, as the forum becomes a platform for members of the

group who may know, or know of, each other to interact online. Instances of this include the

membership of statisticians and users of structural equation modeling packages in SEMNET. While

a list with open membership, the active members of SEMNET are contributors to the field whose

academic publications are widely read and heavy users of packages like LISREL and PLS.

Further, continuance of membership in the online forum in such cases is likely to be linked to

professional or social affiliations, therefore providing a certain level of stability to the membership

of the online group.

August 5 2004 23
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
Groups with stable memberships allow individuals to build up reputations over time by

making contribution to the group in the form of thoughtful comments, providing responses to

requests and providing services to the community such as maintaining FAQ, websites with

supplementary content etc. Participants are also likely to bring online, the social capital they have

offline from contributions to the field and authoring insightful papers. Because of the relative

stability of membership, the visibility payoff from online activities is likely to be considerable. For

instance, it is likely that attending a paid seminar on Structural Equation Modeling by an active

contributor to SEMNET is likely to be seen as being valuable based on the insights contributed by

him or her on SEMNET. This is a phenomenon similar to the payoffs in other professional

contexts for participants in the development of open source software (Lerner and Tirole 2000).

Prior research indicates that subjects asked to read posts on their favorite newsgroup and rate the

authors could not readily associate messages they had just read with authors, even when they

declared themselves to be frequent users of the group (Fiore, Tiernan and Smith 2002). This

suggests that the overlap of membership with a physical group to build social capital through

recognition of the authors identities is likely to be a particularly important issue. It is likely that an

overlap with a physical group can provide supplementary cues to support the creation of a

cumulative memory.

These arguments suggest that social capital plays a role in influencing helping behavior in

contexts of stable membership attributable to the uniqueness of the group and the overlap between

the online group and offline groups. These arguments suggest the following propositions:

P5a: The degree of uniqueness of an online group enhances the relationship between social capital and the level of
helping behavior.
P5b: The degree of overlap between the membership of online and offline groups enhances the relationship between social
capital and the level of helping behavior.

August 5 2004 24
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
4. Discussions and Conclusions

This paper brings together the results of prior work on online groups, and proposes a framework to

study prosocial behavior in online groups. In doing so, we have attempted to explain more fully the

rationale and the contextual factors associated helping behavior and the benefits from participation

in online groups. We highlight the complex effects of the size of the network on the motivation to

contribute. Drawing from the literature on groups, we also highlight the differential impact of two

dimensions of diversity: expertise diversity and cognitive diversity on helping and the usefulness of

online groups. Recognizing information sharing and knowledge sharing as two basic types of

interaction, we highlight how the creation of ancillary resources such as FAQs and repositories of

content supplementing the site change the dynamics of the interaction suppressing information

exchange while enhancing knowledge sharing. Viewing online groups as forums for highly visible

action by participants, we highlight how these differences in the visibility payoffs for information

and knowledge sharing can provide an explanation for the participation of both central and

peripheral participants. In doing so, we highlight the unrecognized role of peripheral participation in

groups and the complementary roles of central and peripheral participants. Finally, we articulate the

conditions under which social capital is likely to be a meaningful influence on behavior in online

groups.

The framework raises several issues for empirical research. For instance, if the creation of

FAQs reduces the incidence of information contribution, this in essence, reduces the opportunity

for peripheral members to perform these basic functions that enable them to develop their

competence and attempt to become more central in the group. Inhibiting the participation of

peripheral members is likely to be counterproductive as it is likely that the ability of the forum to

generate new ideas is impaired. These are all issues with considerable implications for the

understanding of the dynamics of online groups and have practical implications as well.

August 5 2004 25
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
The issues raised in the paper can be empirically examined using data from behavior on newsgroups

and archives of Usenet posting available online at sites such as google.com. Recent efforts to

develop tools (Smith 2002) that provide information on social participation in online groups such as

the level of activity, the size of the persistent posters, the number of discussion topics (threads), the

number of posts on each topic and the behavior of individuals across time are likely to be very

useful in this regard.

The examination of online groups can provide an understanding of similar phenomena in

organizational groups. A variety of firms have established organizational networks to as forums for

employees to collaborate with others in similar functions across the organization and overcome the

barriers of geography in leveraging expertise. The relationships we highlight for online communities

are likely to be applicable to phenomena in organizational networks as well.

August 5 2004 26
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
References

Bandura, A. 1977. Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.

Batson, C. D. 1998. Altruism and prosocial behavior. In D. T. Gilbert, S. Fiske, & G. Lindzey (Eds.),
The handbook of social psychology (4th ed.): 282-316. New York: McGraw-Hill.

Betancourt, H. 1990. An attribution-empathy model of helping behavior: behavioral intentions and


judgments of help-giving. Personality and Social Psychology Bulletin, 16: 573-591.

Brown, J. D., and Smart, S. A. 1991. The self and social conduct: linking self-representations to
prosocial behavior. Journal of Personality and Social Psychology, 60: 368-375.

Butler, B., Sproull, L., Kiesler, S. & Kraut, R. 2002. Community effort in online groups: Who does
the work and why? In S. Weisband & L. Atwater (Eds.), Leadership at a Distance available online
at http://opensource.mit.edu/papers/butler.pdf, accessed January 2, 2004.

Brent, G, Gibbs and M. 1993. Denominational affiliation change: Application of the Consumer
Decision Model. Journal of Consumer Affairs, Winter.

Chaudhuri; A and Holbrook, M. The chain of effects from brand trust and brand affect to brand
performance: The role of brand loyalty, Journal of Marketing; New York; Apr 2001; 65(2) pp. 81-
93

Constant, D., Sproull, L., and Kiesler, S. 1996. The kindness of strangers: The usefulness of
electronic weak ties for technical advice. Organization Science, 7: 119 135.

Cross, R. "More Than an Answer: How Seeking Information Through People Facilitates Knowledge
Creation," Presentation at the Academy of Management Conference, Toronto, 2000.

Daft, R. L. and Lengel, R. H. 1986. Organizational Information Requirements, Media Richness and
Structural Design. Management Science (32,5). 554-571.

Finholt, T. & Sproull, L. 1990. Electronic groups at work. Organization Science, 1: 41-64.

Fiore, Andrew T., Scott Lee Teirnan and Marc A. Smith, Observed behavior and perceived value of
authors in Usenet newsgroups: Bridging the gap, Proceedings of CHI 02, Minneapolis, Apr.20 25,
2002.

Franke, N. & Shah, S. 2001. How community matters for user innovation: The open source of
sports innovation. Sloan Working Paper # 4164.

Granovetter, M. 1973. The strength of weak ties. American Journal of Sociology, 78(6): 1360-
1380.

Hutchins E. 1990. The Technology of Team Navigation. In Galegher J. R., Kraut R. E. and Egido L.
(eds.) Intellectual Teamwork: Social and Technological Foundations of Co-operative Work.
Lawrence Erlbaum Associates,191-220.
August 5 2004 27
Determinants of Helping Behaviors in Online Groups: A Conceptual Model

Kollock, P. 1999. The economies of online cooperation: gifts and public goods in cyberspace. In
Smith, M. A. & Kollock, P. (Eds.), Communities in Cyberspace. Routledge, London, 220-239.

Konstan, J.A., Miller, B.N., Maltz, D., Herlocker, J.L., Gordon, L.R. & Riedl, J.T. 1997. GroupLens:
Applying collaborative filtering to usenet news. Communications of the ACM, 40(3): 77-87.

Latane, B. and Darley, J. 1970. The unresponsive bystander: Why doesnt he help? Englewood
Cliffs, NJ: Prentice Hall.

Lerner, J. & Tirole, J. 2000 (December 29). The simple economics of open source. Retrieved May
10, 2002 from the World Wide Web: http://www.people.hbs.edu/jlerner/simple.pdf

Levine, R. V., Martinez, T. S., Brase, G., & Sorenson, K. (1994). Helping in 36 U.S. cities. Journal of
Personality and Social Psychology, 67: 69-82.

Libby, R., Trotman, K. T., & Zimmer, I. 1987. Member variation, recognition of expertise, and
group performance. Journal of Applied Psychology, 72: 81-97.

Littlepage, G., Robinson, W., & Reddington, K. 1997. Effects of task experience and group
experience on group performance, member ability, and recognition of expertise. Organizational
Behavior and Human Decision Processes, 69(2): 133-147.

Miller, C. C., Burke, L. M., & Glick, W. H. 1998. Cognitive diversity among upper-echelon
executives: Implications for strategic decision processes. Strategic Management Journal, 19: 39-
58.

Nonaka, I. Dynamic Theory of Organizational Knowledge Creation 1994. Organization Science


5(1), pp. 14-37.

Pelled, L.H., Kathleen M Eisenhardt, Katherine R Xin. 1999. Exploring the black box: An analysis
of work group diversity, conflict, and performance. Administrative Science Quarterly. 44(1): 1-28.

Pickering, J. M., & King, J. L. 1992. Hardwiring Weak Ties: Individual and
Institutional Issues in Computer Mediated Communication. Paper presented at the ACM
Conference on Computer-Supported Cooperative Work (CSCW '92), Toronto, Canada.

Rau, D. 2000. Knowing who knows what: The effect of transactive memory on the relationship
between diversity of expertise and performance in top management teams. Working Paper.
University of Minnesota.

Shanteau, James. 1992. Competence in Experts: The Role of Task Characteristics. Organizational
Behavior and Human Decision Processes, 53, 252-266.

Smith, Marc A., 2002. Tools for Navigating Large Social Cyberspaces. Communications of the
ACM, 45(4). 51-55.

August 5 2004 28
Determinants of Helping Behaviors in Online Groups: A Conceptual Model
Sproull, L. & Faraj, S. 1995. Atheism, Sex, and Databases: The net as a social technology. In B.
Kahin & J. Keller (Eds.), Public Access to the Internet: 62-81. Cambridge: MIT Press.

Steblay, N. M. 1987 Helping behavior in rural and urban environments: A meta- analysis.
Psychological Bulletin, 102: 346-356.

Szulanski, Gabriel. 1996. Impediments to the Transfer of Best Practices Within the Firm. Strategic
Management Journal. 17. 27-43.

von Hippel, Eric, 1994. Sticky Information and the Locus of Problem Solving: Implications for
Innovation. Management Science. 40, 4. 429-439.

Wasko, M. M. and Faraj, S. (2000). It is What One Does: Why people participate and help others in
electronic communities of practice. Journal of Strategic Information Systems, 9: 155-173

Wenger, E. 1998. Communities of Practice. Cambridge, UK: Cambridge University Press.

Williams, K.Y. and OReily, C.A. III. 1998. Demography and diversity in organizationa: A review of
40 years of research. In L.L. Cummings and B.M. Staw (Eds.). Research in Organizational
Behavior, 20. 77-140.

Woods, D.D., Patterson. E, Roth. E. M.,Christoffersen, K. 2001. Can we ever escape from data
overload? Working Paper, Ohio State University.

Zhang, W. & Storck, J. 2001. Peripheral members in online communities. Proceedings of the
Americas Conference on Information Systems, Boston, MA.

August 5 2004 29
Determinants of Helping Behaviors in Online Groups: A Conceptual Model

Figure 1: Model of Factors Influencing Helping Behavior in Online Groups

Helping Behaviors
Social Capital

Information Contribution
Knowledge Contribution

Nature of Online group

Size (+/-)
Diversity (+/-)
Ancillary Resources (+/-)
Role of Peripheral and Central
Members (+/-)

Group Characteristics

Uniqueness of forum
Degree of overlap of membership
with membership of Physical Group

Note: the directionality of the associations is indicated in parentheses.

August 5 2004 30

You might also like