25.parental Management of Adolescent Peer Relationships in Context

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Journal of Family Psychology Copyright 2002 by the American Psychological Association, Inc.

2002, Vol. 16, No. 1, 58 – 69 0893-3200/02/$5.00 DOI: 10.1037//0893-3200.16.1.58

Parental Management of Adolescent Peer Relationships in Context:


The Role of Parenting Style
Nina S. Mounts
Northern Illinois University

Three hundred 9th-graders and their best friends participated in this short-term longitudinal
study. Confirmatory factor analysis of the Parental Management of Peers Inventory (N.
Mounts, 2000) suggested that the 4-factor structure provided a good fit to the data. Significant
differences in adolescents’ perceptions of parental management of peers occurred across 4
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

parenting style groups. Structural equation modeling was used to examine the relation among
This document is copyrighted by the American Psychological Association or one of its allied publishers.

5 aspects of parental management of peers and adolescents’ Time 1 drug use, friends’ Time
1 drug use, and adolescents’ Time 2 drug use. The parental management styles of monitoring,
guiding, prohibiting, and supporting all had significant paths in the model, whereas neutrality
was not significant. Multiple group comparisons were used to examine whether parenting
style moderates the relation between parental management of peers and the drug use
outcomes. Parenting style functioned as a moderator for 7 of the paths in the model.

Drawing on Bronfenbrenner’s (1979) ecological theory, Several researchers have suggested a distinction between
researchers have devoted much time to examining the link- parenting style and parenting practices (Darling & Stein-
ages between parent and peer relationships. Much of the berg, 1993; Mize, Russell, & Pettit, 1998). According to
research on linkages between parent and peer contexts has these researchers, parenting style is characterized by the
focused on parenting and social competence in younger attitudes toward the child and the emotional climate of the
children (Parke & Ladd, 1992). Because of changes that parent– child relationship rather than specific, goal-directed
occur in both parent– child and peer relationships during behavior. In contrast, parenting practices have specific goals
adolescence, it is interesting to consider the ways in which and specific content. That is, when parents use parenting
adolescents’ relationships with parents may affect their re- practices, their focus is on affecting a particular aspect of
lationships with peers. During adolescence, there is an in- the child. Parke and O’Neil (1999) suggested a tripartite
crease in susceptibility to peer influence (Berndt, 1979) model in which parenting style is one of the facets of
concomitant with an increased interest in peer relationships parenting that affects peer relationships. In this model,
more generally (Savin-Williams & Berndt, 1990). At the parenting practices are subdivided into those practices in
same time, there is a decline in susceptibility to parental which parents function as a direct instructor in regard to
influence (Steinberg & Silverberg, 1986). In addition, nu- peer relationships and those practices in which parents
merous investigators have found that peer influence is a function to provide children with opportunities for peer
strong predictor of adolescent substance use (Hawkins, interactions.
Catalano, & Miller, 1992; Urberg, Cheng, & Shyu, 1991), The literature that explores the ways in which parent and
premarital sexual activity (Billy & Udry, 1985), delinquent peer contexts are linked during adolescence is not extensive.
behavior (Magnusson, Stattin, & Allen, 1986), and that it is However, some research exists that focuses on parenting
also related to school adjustment (Berndt, 1999). Because of style and its relation to peers (Fuligni & Eccles, 1993;
this, the nature of this linkage between parent and peer Mounts & Steinberg, 1995). Fuligni and Eccles found that
contexts is of interest. when adolescents believed that their parents were restrictive
Recently, there has been recognition in the literature that and asserted too much power, they reported higher levels of
there are several facets to parental influences on children. peer orientation. Similarly, Mounts and Steinberg reported
that when adolescents reported that their parents used a
nonauthoritative style, they were more likely to be influ-
Support for this study was provided by B/START Grant enced by their peers to use drugs.
MH57232-01 from the National Institute of Mental Health. I Although a number of studies have examined parenting
acknowledge David Valentiner for his advice on the data analyses. practices associated with young children’s peer relation-
Karen Bogenschneider provided helpful comments on drafts of the ships (Ladd & Golter, 1988; Mize & Pettit, 1997), the
article. Penny Brucker, Jennifer Marchand, Benjamin Perry, and literature that examines parenting practices associated with
Steven Toefl conducted the data collection. I am especially grateful
to the children, parents, and schools for their participation in this adolescents’ peer relationships is quite limited. There is a
project. considerable amount of research that examines parental
Correspondence concerning this article should be addressed to monitoring during adolescence (Bogenschneider, Raffaelli,
Nina S. Mounts, Department of Psychology, Northern Illinois & Tsay, 1998; Stattin & Kerr, 2000; Steinberg, Lamborn,
University, DeKalb, Illinois 60115. E-mail: nmounts@niu.edu Darling, Mounts, & Dornbusch, 1994). Because parents

58
PARENTAL MANAGEMENT OF ADOLESCENT PEER RELATIONSHIPS 59

enable or restrict access to peers by monitoring, monitoring at home where adolescents can have their friends over.
is considered to be a parenting practice with which parents These strategies were the focus of the present investigation.
manage peer relationships (Parke & O’Neil, 1999). There is Darling and Steinberg (1993) suggested that to move the
strong support in the literature for the relation between socialization literature forward, researchers must begin to
parental monitoring and involvement with deviant peers consider both parenting style and parenting practices jointly
during adolescence (Brown, Mounts, Lamborn, & Stein- when examining their effects on children. As such, they
berg, 1993). Several studies reported a significant relation suggested that the way in which parenting style affects
between low levels of parental monitoring and delinquent children’s development is through the attitudes that the
child behavior. Steinberg (1986) found that adolescents in parents hold in regard to the child. The parenting style that
self-care after school who were monitored distally by their parents have provides a context within which the parenting
parents, through telephone calls home, were less susceptible practices are made more or less effective. In short, they
to antisocial peer pressure than were adolescents whose suggested that parenting style moderates the effects that
parents did not monitor them. Similarly, Fuligni and Eccles parenting practices have on children’s development. Few
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

(1993) found that high levels of monitoring were associated empirical studies exist that examine the joint contributions
This document is copyrighted by the American Psychological Association or one of its allied publishers.

with a less extreme peer orientation. In contrast, Bogen- of parenting style and parenting practices to child or ado-
schneider et al. (1998) reported no significant relation be- lescent development. One study that examined the ways in
tween monitoring and peer orientation. These mixed results which parenting style functions as a moderator comes from
in regard to monitoring at least suggest that there are other the work of Steinberg, Lamborn, Dornbusch, and Darling
variables that may act to moderate the relation between (1992). They found that when parents were nonauthorita-
monitoring and peer relationships. Beyond research that tive, it attenuated the effects of parental school involvement
examines monitoring, there is very little research that ex- in determining school performance, whereas parental in-
amines the ways in which parents affect the peer relation- volvement in school had an even stronger effect in the
ships of adolescents. presence of the authoritative parenting style.
Vernberg, Beery, Ewell, and Abwender (1993) conducted Mize and Pettit (1997) examined the ways in which
one study that did examine parenting practices, other than parenting style and parenting practices affect children’s
monitoring, associated with adolescents’ peer relationships. social competence. Using a sample of 3- to 6-year-old
children, Mize and Pettit’s results partially supported the
Four groups of friendship facilitation strategies were iden-
moderator model suggested by Darling and Steinberg
tified as being useful in helping seventh- and eighth-graders
(1993). Contrary to expectation and to the work of Steinberg
form friendships after a move to a new school district.
et al. (1992), effective parenting styles did not operate to
Parents reported that they met with other parents, facilitated
make parenting practices even more effective. Rather, Mize
proximity to peers, talked with their adolescent about peer
and Pettit’s work suggested that the positive parenting prac-
relationships, and encouraged activities with other adoles- tices became even more important in the absence of more
cents as strategies for helping their young adolescent make positive parenting styles. When mothers were low in their
new friends after a family relocation. Variability across levels of general parental responsiveness, constructive
families in the frequency of use of these strategies in the coaching in peer relationships was related to lower levels of
first 3 months following a family relocation was found. aggression than when mothers had low levels of responsive-
Adolescents whose parents reported the highest frequency ness and did not use constructive coaching. In contrast,
of friendship facilitation strategies had greater success in coaching did not affect aggression when mothers had mod-
making friends. erate or high levels of responsiveness toward their children.
In a study conducted by Youniss, DeSantis, and Hender- Steinberg et al.’s (1992) and Mize and Pettit’s studies have
son (1992), hypothetical situations were used to assess the served as an important starting point for further investiga-
number of parent behaviors that would be directed toward tion of the joint contributions of parenting style and parent-
the adolescent in a hypothetical situation involving peers. A ing practices to development. In the present investigation,
modest number of parent behaviors directed toward man- the focus was on adolescents’ perceptions of parental man-
agement of peer relationships was recorded using this agement of peer relationships and parenting style and their
method. association with adolescents’ Time 1 and Time 2 drug use
In another study of parenting practices associated with and friends’ Time 1 drug use.
adolescents’ peer relationships, Mounts (2000) identified a Five questions were of interest in the present investiga-
number of strategies used by parents to affect adolescents’ tion. First, is the four-factor structure of the Parental Man-
peer relationships, including guiding, which is when parents agement of Peers Inventory (PMPI; Mounts, 2000) con-
talk with the adolescent about the consequences of being firmed in this sample? The four-factor structure used in
friends with particular people. Neutrality is when parents do previous investigations (Mounts, 2000, 2001) was subject to
not interfere with their children’s peer relationships, per- confirmatory factor analysis. Second, are the parenting style
haps because of the parents’ belief that who their child variables and the parenting practices variables distinct from
selects as friends is their own individual choice. Prohibiting one another? Several researchers suggested that these par-
is when parents let adolescents know that they do not want enting variables are conceptually distinct from one another
them to associate with particular peers. Finally, supporting (Darling & Steinberg, 1993; Mize et al., 1998; Parke &
is when parents do things such as providing an environment O’Neil, 1999). Before analyzing the moderator relation be-
60 MOUNTS

tween the parenting style and parenting practices variables, parenting style contexts? As suggested by Darling and
it is important to establish that they are, in fact, empirically Steinberg (1993), one would expect different types of man-
distinct from one another. agement to be more effective within some parenting style
Third, are there differences in adolescents’ perceptions of contexts than others. This question assesses processes that
peer management across authoritative, authoritarian, indul- occur within the parenting style context rather than simply
gent, and uninvolved parenting style groups? It may be examining the mean differences across the four parenting
expected that some parenting practices would be used more style groups that were the focus of the third question in this
within certain parenting style contexts and less within other investigation. As such, there could be mean differences in
parenting style contexts. I hypothesized that adolescents management across the parenting style groups, but the re-
who reported that their parents used the authoritative style lation between management and the drug use outcomes may
would report lower levels of prohibiting; higher levels of be very similar across the four groups. Alternatively, mean
monitoring, guiding, and supporting; and moderate levels of differences in management may emerge, and the relation
neutrality. In contrast, I hypothesized that adolescents who between management and drug use outcomes may be dif-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

reported that their parents used the authoritarian style would ferent across the groups.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

report higher levels of monitoring and prohibiting and lower Similar to Steinberg et al. (1992), I hypothesized that the
levels of guiding, neutrality, and supporting. I also hypoth- negative relation between adolescents’ perceptions of mon-
esized that adolescents who reported that their parents used itoring and adolescents’ perceptions of guiding and the
the indulgent parenting style would report higher levels of outcome variables would be enhanced by the authoritative
neutrality and supporting and lower levels of monitoring, style. For adolescents’ perceptions of parental neutrality in
guiding, and prohibiting. Finally, I hypothesized that ado- regard to peer relationships, I hypothesized that higher
lescents who reported that their parents used the uninvolved levels of neutrality within the authoritative context would
style would report the highest levels of neutrality in con- not be associated with higher levels of Time 1 and Time 2
junction with the lowest levels of monitoring, guiding, drug use for adolescents. Further, I hypothesized that higher
prohibiting, and supporting. levels of neutrality would not be associated with having
Fourth, in what ways are adolescents’ perceptions of friends with higher levels of Time 1 drug use. However, I
parental management practices related to adolescents’ Time hypothesized that adolescents’ perceptions of higher levels
1 drug use, adolescents’ Time 2 drug use, and friends’ Time of neutrality within the indulgent or uninvolved contexts
1 drug use? Figure 1 presents the hypothesized relation would be associated with higher levels of adolescents’ Time
between the parenting practices variables, adolescents’ 1 and Time 2 drug use as well as with having friends with
Time 1 drug use, adolescents’ Time 2 drug use, and friends’ higher levels of Time 1 drug use. I hypothesized that higher
Time 1 drug use. levels of prohibiting within the context of the authoritative
Fifth, does the relation between parental management of or indulgent styles would have minimal effects, whereas
peers and the drug use outcomes vary within different higher levels of prohibiting within the authoritarian style

Figure 1. Hypothesized structural equation model of relation between parental management


strategies and drug use.
PARENTAL MANAGEMENT OF ADOLESCENT PEER RELATIONSHIPS 61

would have the most negative effects on adolescents be- had the entire class period (approximately 45 min) to complete the
cause the context of the authoritarian relationship is already questionnaire. Graduate students who were members of the re-
quite restrictive. Finally, as suggested by the work of Mize search team circulated through the classrooms to answer any
and Pettit (1997), I hypothesized that adolescents’ percep- questions, to check to see that identification numbers were re-
corded accurately, and to collect the completed materials at the end
tions of higher levels of support may offset some of the of the class period. Identification numbers were used to match
negative effects of the authoritarian and uninvolved styles. adolescent questionnaires across all waves of data collection. To
maximize the sample size, research assistants returned to the
Method school on multiple occasions to collect data from adolescents who
were not available to complete questionnaires when the main data
Sample collection occurred.
The 300 participants included in these analyses were drawn
from a larger, 1-year longitudinal study of parental influences on Measures
ninth-graders’ peer relationships. The school was located in a Parenting style. Adolescents’ perceptions of parental warmth
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

small, midwestern city. All ninth-graders present in the school on and of behavioral control were assessed using the 56-item revision
This document is copyrighted by the American Psychological Association or one of its allied publishers.

the testing days were asked to complete questionnaires. Question- of the Child’s Report of Parent Behavior Inventory (CRPBI;
naire data were collected from adolescents and their friends in two Burger & Armentrout, 1971; Schaefer, 1965). Adolescents re-
waves (beginning of the school year and end of the school year). sponded to items about their parent such as “Gives me a lot of care
Data from friends were matched with target ninth-graders’ data and attention” (Warmth subscale) and “Does not insist I obey
and used in the analyses. if I complain or protest” (Behavioral Control subscale [reverse
The demographics of the sample were as follows: Approxi- coded]). Confirmatory factor analysis was conducted on these
mately 40% of the sample were boys and 60% of the sample were items to examine the two-factor structure reported in previous
girls. The family structure of the adolescents was as follows: intact investigations. Not surprisingly, factor loadings were all signifi-
(58%), single parent (21%), stepfamily (17.3%), and other (3.7%). cant, and the analyses suggested a good fit to the data, ␹2(701, N ⫽
The majority of the sample was White (83.8%); the remainder of 300) ⫽ 1,272.77; normed fit index (NFI) ⫽ .96; comparative fit
the sample was composed of African Americans (8.2%), Asians index (CFI) ⫽ .98; relative fit index (RFI) ⫽ .96; root-mean-
(3.2%), Hispanics (1.6%), Native Americans (1.6%), and others square error of approximation (RMSEA) ⫽ .05. Bentler and Bon-
(1.6%). Father’s education was represented by less than a high nett (1980) suggested that a NFI greater than .90 suggests a good
school education (17%), high school graduate (35%), trade school fit. The closer the CFI is to 1, the better the fit (Bentler, 1990). RFI
or some college (27%), college graduate (14.1%), and graduate values close to 1 indicate a very good fit (Arbuckle & Wothke,
degree (6.9%). 1999).
Factor scores were used to compute the parental warmth and the
Procedure behavioral control variables. Median splits on the parental warmth
and behavioral control variables were then conducted to create the
Because the investigation involved questions regarding adoles- four parenting style groups. Adolescents who reported high paren-
cents and their best school friends, there was concern that a tal warmth and high behavioral control were considered to be from
standard active consent procedure (in which parents of the adoles- authoritative homes, those who reported low warmth and high
cents and the parents of the adolescents’ friends return signed behavioral control were considered to be from authoritarian
consent forms) would produce a very small sample. Therefore, I homes, those who reported high warmth and low behavioral con-
decided to use a passive consent procedure in this investigation. In trol were said to be from indulgent homes, and those who reported
consultation with the administrators of the participating school and low warmth and low behavioral control were said to be from
the university’s Institutional Review Board, a consent procedure uninvolved homes. Using this method, I placed 27.2% of the
was used in which assent was solicited from the adolescents and adolescents in the authoritative parenting style group, 22.5% in the
their friends, whereas passive consent was sought from the parents. authoritarian parenting style group, 22.8% in the indulgent parent-
All parents in the school were mailed a letter, via first-class mail, ing style group, and the remaining 27.5% in the uninvolved par-
approximately 4 weeks prior to the first scheduled data collection. enting style group.1
Parents were asked to call either the school or the research office Parenting practices. Adolescents’ perceptions of parental
if they did not want their child to participate in the research study. monitoring were assessed using a five-item scale on which ado-
Approximately 1% of the adolescents in the school had consent lescents indicated the degree to which their activities were moni-
withheld. tored by their parents (Dornbusch et al., 1985). Adolescents re-
Assent was solicited from all students on the day of the first data
collection. Students were provided a written description of the
1
study, including their role in the study and the voluntary nature of An alternate strategy for creating the parenting style groups
their participation. Students who wished to participate signed the was suggested by one of the reviewers, and analyses were con-
form and then proceeded to respond to the questionnaire. Students ducted to ascertain similarity in groupings across the two methods.
who did not wish to participate in the investigation were allowed In the alternate method, factor scores were derived from an addi-
to work on school-related activities at their desks or were allowed tional exploratory analysis that was conducted on the single factors
to go to the library. No adolescents refused to participate in the (parental warmth and behavioral control) that were tested in the
research study. two-factor confirmatory factor analysis. Median splits were then
Data were collected using self-report questionnaires adminis- conducted on the factor scores to create the parenting style groups.
tered to adolescents in classroom-size groups (about 25 to 30 The strategy described in the text and this alternate strategy
students) during a required social studies class. Research staff yielded comparable results. Across both methods of parenting style
distributed the questionnaires and instruction sheets and read the group formation, approximately 98% of the adolescents were in
instructions aloud at the beginning of the class period. Students the same parenting style group.
62 MOUNTS

Table 1
Means, Standard Deviations, and Intercorrelations Among the Major Variables
Variable M SD 2 3 4 5 6 7 8
1. Monitoring 2.28 .54 .14* .13* ⫺.11 .22*** ⫺.37*** ⫺.26*** ⫺.36***
2. Guiding 1.02 .25 — .09 .49*** .50*** .01 ⫺.17** ⫺.12
3. Neutrality 1.41 .36 — ⫺.16** .29*** ⫺.07 ⫺.08 ⫺.10
4. Prohibiting 1.50 .41 — .25*** .18*** .13* .15*
5. Supporting 1.28 .34 — ⫺.14* ⫺.19** ⫺.15*
6. Adolescents’ Time 1 drug use 1.78 .88 — .52*** .74***
7. Friends’ Time 1 drug use 1.68 .70 — .54***
8. Adolescents’ Time 2 drug use 1.70 .83 —
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

sponded to items such as “How much do your parents really know Friends’ Time 1 drug use was calculated by first calculating
This document is copyrighted by the American Psychological Association or one of its allied publishers.

where you are at night?” Adolescents used a three-point scale individual friend’s Time 1 drug use scores. Then, the Time 1 drug
ranging from 1 (don’t really know) to 3 (know a lot). Alpha for the use scores of the friends were averaged. At least two friends had
scale was .82. A high score on the scale indicates a high level of to complete the Time 1 drug use measure for them to be included
parental monitoring. in the analyses. These scores were then entered as the friends’
Adolescents’ perceptions of direct parental involvement in peer Time 1 drug use score in all analyses.
relationships were assessed using four scales derived from the
25-item PMPI (Mounts, 2000). A four-point Likert-type scale was Results
used on which responses ranged from 1 (strongly disagree) to 4
(strongly agree). Perceptions of parental guiding were assessed Means and standard deviations for the major variables as
with nine items; adolescents were asked to respond to items such well as correlations among the major variables are presented
as “My parents talk to me about the pros and cons of hanging in Table 1.
around with certain people.” Perceptions of parental neutrality
were assessed with five items; adolescents responded to items such
as “My parents tell me that who I have as friends is my personal
Confirmatory Factor Analysis
choice.” Perceptions of parental prohibiting of peer relationships Because the PMPI has been used in previous investiga-
were assessed with six items that demonstrated the extent to which tions but has not been examined using confirmatory factor
parents attempted to limit adolescents’ interactions with particular analyses, I used confirmatory factor analysis to examine the
peers, for example, “My parents tell me if they don’t want me to
hang around with certain kids.” A high score on this scale indicates
factor structure of the scale. The confirmatory factor anal-
a high level of prohibiting of friendships. Perceptions of parental ysis suggested a good fit to the data, ␹2(269, N ⫽ 300) ⫽
supporting of peer relationships was assessed using five items; 820.35; NFI ⫽ .96; CFI ⫽ .97; RFI ⫽ .95; RMSEA ⫽ .07.
adolescents were asked to respond to items such as “My parents Results are presented in Table 2. Factor loadings were all
encourage me to invite kids they like over to my house.” significant. The NFI was .96, suggesting a good fit to the
Friendship nominations. As suggested by Kandel (1978), ad- data. Similarly, the CFI was .97, indicating a good fit to the
olescents were asked to name their three best friends in the same data. The correlations among the four factors were as fol-
school and grade. This information was used to match adolescents lows: guiding and neutrality (r ⫽ .15, p ⬍ .05), guiding and
and their friends to obtain measures of friends’ Time 1 drug use. prohibiting (r ⫽ .59, p ⬍ .001), guiding and supporting (r ⫽
Because adolescents may inflate the degree of similarity between .72, p ⬍ .001), neutrality and prohibiting (r ⫽ ⫺.25, p ⬍
themselves and their friends, having friends report on their own
drug use and using those measures of drug use in analyses sub-
.001), neutrality and supporting (r ⫽ .41, p ⬍ .001), and
stantially increases the validity of the assessment (Bauman & prohibiting and supporting (r ⫽ .34, p ⬍ .001).2 Factor
Fisher, 1986). scores from the confirmatory factor analysis were used to
Adolescents’ and friends’ drug use. Adolescents and their best create the four parental management variables.
friends were asked to report “How often in the last three months
have you” used various types of drugs. McCord (1990) suggested Correlations Between the Parenting Style Variables
that although self-reports of antisocial behavior are subject to and Parental Management of Peers
under- and overreporting, they appear to be more valid methods of
assessing problem behaviors in adolescence than are police Because the focus of this investigation was on adoles-
records. Furthermore, they are more valid than observational meth- cents’ perceptions of two different aspects of parenting, it
ods because many antisocial behaviors are not readily observable.
Participants completed this measure at Time 1 (beginning of the
2
school year) and at Time 2 (end of the school year). Friends Because the correlation between guiding and supporting was
completed this measure at Time 1. Four response options were .72, at the suggestion of one of the reviewers, I conducted an
given ranging from Never to Often. A high score on these scales additional confirmatory factor analysis in which guiding and sup-
indicates high levels of drug use. Drug use was assessed using a porting were combined into one factor rather than two separate
five-item scale that assesses students’ use of alcohol, marijuana, factors. The chi-square difference between the two models sug-
and other drugs (Lamborn, Mounts, Steinberg, & Dornbusch, gested that the model in which the factors were separate had a
1991). Alphas for this scale were .92 and .93 at Time 1 and Time better fit to the data than did the model in which the guiding and
2, respectively. supporting factors were combined.
PARENTAL MANAGEMENT OF ADOLESCENT PEER RELATIONSHIPS 63

Table 2
Factor Loadings From the Confirmatory Factor Analysis of Items on the Parental Management of Peers Inventory
Factor loading Item
Guiding friendships
.28 My parents talk to me about the pros and cons of hanging around with certain people.
.55 My parents tell me that who I have for friends will affect my future.
.56 My parents want me to be friends with kids who don’t drink alcohol or do drugs.
.55 I’m only friends with kids who are good students because my parents want me to be.
.60 I’m only friends with kids who don’t drink or do drugs because my parents want me to be.
.57 My parents only want me hanging around with kids who are like our family.
.48 If my friends do things that my parents don’t approve of I stop being friends with them.
.38 My parents influence my selection of friends.
.42 My parents try to be in charge of my friendships.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Neutrality
This document is copyrighted by the American Psychological Association or one of its allied publishers.

.66 My parents tell me that who I have as friends is my personal choice.


.65 My parents don’t interfere with my friendships.
.46 My parents do not concern themselves with my friendships.
.69 My parents think that who I have for friends is my own business.
.40 My parents don’t talk to me about my friends.
Prohibiting friendships
.69 My parents tell me that they don’t like my friends.
.70 My parents tell me if they don’t want me to hang around with certain kids.
.70 My parents tell me that they don’t approve of the things my friends do.
.42 My parents want me to be friends with kids who are good students.
.62 My parents let me know who they want to be my friends.
.53 My parents think that if my friends are doing bad things, I must be doing them too.
Supporting friendships
.54 My parents encourage me to invite kids they like over to my house.
.56 My parents encourage me to do activities with kids they like.
.52 My parents help me think of ways to meet new kids.
.54 My parents want me in certain activities at school because of the kinds of kids I’ll meet in them.
.54 My parents support me in my activities because they like the friends I meet in them.

was important to establish that the constructs were, in fact, dent variable in the analyses. No significant gender differ-
distinct. High correlations between the parenting style di- ences were found, nor were there significant Parenting
mensions and the parenting practices were consistent with Style ⫻ Gender interactions. Thus, I conducted a reanalysis
the position that the constructs were not distinct. Such a with gender dropped from the model. To control for fami-
finding would make it questionable to conduct analyses that lywise error associated with conducting multiple ANOVAs,
examine parenting style as a moderator of parental manage- a significance level of p ⬍ .01 was set for the F values
ment of peers. I computed the correlations of the parenting associated with the five ANOVAs. I conducted Bonferroni
style variables with the parental management variables by tests to assess differences between pairs of parenting style
conducting a confirmatory factor analysis on the parenting groups. Results of these analyses are presented in Table 3.
style variables and the parental management variables. The I found significant differences in adolescents’ perceptions
range in the absolute value of the correlations was between of management across the four parenting style groups for
.02 and .39, with the mean of the absolute values being .17. monitoring, F(3, 296) ⫽ 10.71, p ⬍ .001; neutrality, F(3,
This suggested that the constructs were distinct and that 296) ⫽ 5.37, p ⬍ .001; and supporting, F(3, 296) ⫽ 4.43,
further examination of the moderator effect was appropriate. p ⬍ .005. I found no differences across the parenting style
groups for adolescents’ perceptions of guiding, F(3, 296) ⫽
Parenting Style Group Differences in Parental .37, p ⫽ ns, or prohibiting, F(3, 296) ⫽ .57, p ⫽ ns. For
Management of Peers monitoring, adolescents from authoritative homes reported
significantly higher levels of monitoring than adolescents
To examine differences in adolescents’ perceptions of from authoritarian, indulgent, and uninvolved homes re-
peer management strategies (monitoring, guiding, neutral- ported. Adolescents who were from authoritative homes and
ity, prohibiting, and supporting) across the four parenting indulgent homes reported significantly higher levels of neu-
style groups (authoritative, authoritarian, indulgent, and un- trality than did adolescents from authoritarian homes. Fi-
involved), I conducted five separate analyses of variance nally, adolescents from authoritative homes reported signif-
(ANOVAs). Because management strategies may vary icantly higher levels of supporting than adolescents from
across gender groups, I also included gender as an indepen- authoritarian or uninvolved homes reported.
64 MOUNTS

Table 3
Means and Standard Deviations of the Parenting Practices Within the Four Parenting Style Groups
Authoritative Authoritarian Indulgent Uninvolved
Variables M SD M SD M SD M SD
Monitoring 2.54a,b,c 0.42 2.18a 0.53 2.32b 0.54 2.14c 0.53
Guiding 1.00 0.19 1.00 0.26 1.04 0.24 1.02 0.30
Neutrality 1.49a 0.29 1.29a,b 0.45 1.50b 0.28 1.38 0.37
Prohibiting 1.51 0.34 1.52 0.46 1.48 0.34 1.44 0.45
Supporting 1.38a,b 0.30 1.23a 0.34 1.33 0.32 1.22b 0.36
Note. Means with subscripts that are the same within a row are significantly different from one another at p ⬍ .05.

Relation Between Parental Management of Peers When adolescents reported that their parents used higher
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

and Drug Use Outcomes levels of prohibiting, they reported higher levels of Time 1
This document is copyrighted by the American Psychological Association or one of its allied publishers.

drug use and associated with friends who reported higher


Structural equation modeling was used to examine the levels of Time 1 drug use. Adolescents who reported that
relation between adolescents’ perceptions of parental man- their parents used higher levels of supporting reported lower
agement and the Time 1 and Time 2 outcomes. It should be levels of Time 1 drug use. As one would expect, adolescents
noted that none of the variables included in this set of who reported higher levels of Time 1 drug use associated
analyses or the moderator effect analyses were latent vari- with friends with higher levels of Time 1 drug use. Simi-
ables. Therefore, the model being tested was not a structural larly, adolescents who reported higher levels of Time 1 drug
model but a path model that was being tested using some of use reported higher levels of Time 2 drug use as well.
the procedures available in structural equation modeling Finally, associating with friends with higher levels of Time
statistical programs. In particular, testing the moderator 1 drug use was related to higher levels of adolescents’ Time
effect using traditional path-analytic multiple regression 2 drug use, even after controlling for the adolescent’s own
procedures would be difficult given the categorical nature of Time 1 drug use.
the parenting style variable. According to Glasgow, Dornbusch, Troyer, Steinberg,
The model presented in Figure 1 was used to test the and Ritter (1997), using the adolescents’ Time 1 drug use as
relation between the five Time 1 parenting practices, ado- a control in the analysis limits the amount of variance that
lescents’ Time 1 drug use, adolescents’ Time 2 drug use, other predictors in the equation can explain. Not including
and friends’ Time 1 drug use using structural equation the adolescents’ Time 1 drug use, however, may have
modeling. Some aspects of the model have been tested in a placed more significance on the remaining predictors than
previous investigation (Mounts, 2001); however, the entire was appropriate. This procedure enabled a closer look at the
model has not been examined. These analyses were a test of relation between the parenting variables and changes in
a main effect model wherein all of the parental management drug use over a short time period and was a more conser-
variables were predictors in the model. The purpose of this vative test of the relation between the parenting variables
set of analyses was to allow a comparison between the and the adolescents’ Time 2 drug use. Using this procedure
parameters in the baseline model and the parameters in the of controlling for adolescents’ Time 1 drug use, these results
interaction model. To facilitate comparison between the suggest that parenting practices may affect changes in ad-
models, path coefficients for this baseline model as well as olescents’ drug use over time and may affect associations
the interaction model are presented in Table 4. The path with friends despite the fact that the data are correlational in
coefficients for the baseline model are reported in the first nature.
column of Table 4.
Adolescents’ perceptions of parental monitoring, guiding, Parenting Style as a Moderator of the Parental
neutrality, prohibiting, and supporting were treated as ex- Management of Peers’ Effects
ogenous variables. The endogenous Time 1 variables were
adolescents’ drug use and friends’ drug use, and the endog- In the final set of analyses, I examined the moderating
enous Time 2 variable was adolescents’ drug use. As illus- effect of parenting style on management practices using
trated in Figure 1, causal paths were permitted between the structural equation modeling. In this set of analyses, a
exogenous variables and the Time 1 and Time 2 endogenous multiple-group structural equation model approach using
variables. Pathways were also permitted between the Time Analysis of Moment Structures (AMOS; Arbuckle &
1 and Time 2 endogenous variables. In this fully saturated Wothke, 1999) was used to test for differences in the
model, all variables were allowed to freely correlate with relationships in Figure 1 across each of the four parenting
one another, ␹2(0, N ⫽ 300) ⫽ 0.00. style groups. Parker, Baltes, and Christiansen (1997) sug-
As can be seen in Table 4, adolescents who reported that gested that several steps are necessary to test models across
their parents used higher levels of monitoring reported groups. In the first step, the fully variant model was esti-
lower levels of Time 1 drug use. Adolescents who reported mated to generate the path coefficients between the eight
that their parents used higher levels of guiding associated constructs (monitoring, guiding, neutrality, prohibiting, sup-
with friends who reported lower levels of Time 1 drug use. porting, adolescents’ Time 1 drug use, friends’ Time 1 drug
PARENTAL MANAGEMENT OF ADOLESCENT PEER RELATIONSHIPS 65

use, and adolescents’ Time 2 drug use) for each of the four

11.18***
11.65**

14.96**
11.34*

7.63*

8.75*

8.94*
groups. These coefficients are presented in the second
␹2

0.22
2.02

0.26

0.93

1.07
4.94
6.03

0.31
3.40

2.42
2.75
through fifth columns of Table 4, with each column pre-
senting data for one of the four parenting style groups.
Because the fully variant model was saturated, the model
provided a perfect fit to the data, ␹2(0, N ⫽ 288) ⫽ 0.00. In
Uninvolved

.98*a,b,c

a,b,c
.75***

.44***
.77***
addition, the model for each of the parenting style groups

⫺.98**
a,b

d
.33b

.22b

.06b
for the unconstrained model was fully saturated, ␹2(0, N ⫽
.09

⫺.43

.27
⫺.33

.19

⫺.09
.16

.06
⫺.53

.12
288) ⫽ 0.00.
Next, using an omnibus test, I examined the relationships
between the eight constructs to assess whether they were
Indulgent

equivalent across the four parenting style groups. In this

.42***
.59***
.39***
b,c
⫺.47*

.72*

.36*a

.16b
step, all paths in the model were held invariant across the
⫺.77c

.04c
.25

⫺.63

.07

.16
⫺.03

.14
⫺.08
⫺.13

.10
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

four parenting style groups. A significant chi-square differ-


This document is copyrighted by the American Psychological Association or one of its allied publishers.

ence between the fully variant model and the invariant


model for these analyses would demonstrate that the con-
structs did not function equivalently across the four groups.
Authoritarian
⫺.68***

.54***

.87***
.13a,c,d

A nonsignificant chi-square difference would have indicated


⫺.45*a,b

a,b
.59*
⫺.55b

⫺.27b
.11a

that the constructs functioned equivalently. Because the


.06

⫺.52

.08
⫺.01
⫺.33

.51
⫺.01
⫺.18
⫺.25

fully variant model was saturated, ␹2(0, N ⫽ 288) ⫽ 0.00,


Note. Coefficients in the same row with the same subscripts are significantly different from each other at p ⬍ .05.

the chi-square for the invariant model was equal to the


chi-square difference. The chi-square difference for the
fully invariant model was significant, ␹2(54, N ⫽ 288) ⫽
Authoritative

76.90, p ⬍ .05, and showed a significant decrease in model


.64***

.94***
a,b
.31**
⫺.48**
a,b

fit. This result indicates that the model did not function
⫺.86*a

a
.56a

⫺.17a
.16

⫺.35
.13
.33
.02
⫺.13

⫺.20

⫺.26

.08
.20
.13

equivalently across the four parenting style groups, suggest-


ing that the model differed within the parenting style
contexts.
In the next step, each individual path in the model was
Main effect model

held invariant while the remaining 17 paths were free to


vary. This procedure assessed whether each path functioned
Path Coefficients and Comparisons Within the Four Parenting Style Groups

.27***
⫺.55***

.40***
.64***
⫺.48**
.35**

similarly across the four parenting style groups. A signifi-


.23*
⫺.34*

⫺.01
.07
.07

⫺.03
⫺.02

⫺.08
⫺.16
⫺.05

.18
.12

cant chi-square for a path indicated that the path was not
equivalent across the four parenting style groups. These
analyses were used to identify the ways in which the models
differed across the four parenting style contexts. The results
of these chi-square analyses are presented in the last column
of Table 4. Of the 18 paths tested, 7 of them had a signif-
Adolescents’ Time 1 drug use 3 Adolescents’ Time 2 drug use

icant chi-squares when the path was held invariant across


Adolescents’ Time 1 drug use 3 Friends’ Time 1 drug use

Friends’ Time 1 drug use 3 Adolescents’ Time 2 drug use

the four groups. These paths were guiding to adolescents’


Time 1 drug use, ␹2(3, N ⫽ 288) ⫽ 11.34, p ⬍ .05;
monitoring to friends’ Time 1 drug use, ␹2(3, N ⫽ 288) ⫽
11.65, p ⬍ .01; neutrality to friends’ Time 1 drug use, ␹2(3,
N ⫽ 288) ⫽ 7.63, p ⬍ .05; guiding to adolescents’ Time 2
Monitoring 3 Adolescents’ Time 1 drug use

Monitoring 3 Adolescents’ Time 2 drug use


Prohibiting 3 Adolescents’ Time 1 drug use

Prohibiting 3 Adolescents’ Time 2 drug use


Supporting 3 Adolescents’ Time 2 drug use
Supporting 3 Adolescents’ Time 1 drug use

drug use ␹2(3, N ⫽ 288) ⫽ 8.75, p ⬍ .05; prohibiting to


Neutrality 3 Adolescents’ Time 1 drug use

Neutrality 3 Adolescents’ Time 2 drug use


Guiding 3 Adolescents’ Time 1 drug use

Guiding 3 Adolescents’ Time 2 drug use

adolescents’ Time 2 drug use, ␹2(3, N ⫽ 288) ⫽ 11.18, p ⬍


* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
Monitoring 3 Friends’ Time 1 drug use

Prohibiting 3 Friends’ Time 1 drug use


Supporting 3 Friends’ Time 1 drug use
Neutrality 3 Friends’ Time 1 drug use

.05; adolescents’ Time 1 drug use to adolescents’ Time 2


Guiding 3 Friends’ Time 1 drug use

drug use, ␹2(3, N ⫽ 288) ⫽ 14.96, p ⬍ .01; and friends’


Path

Time 1 drug use to adolescents’ Time 2 drug use, ␹2(3, N ⫽


288) ⫽ 8.94, p ⬍ .05. These results provide evidence that
these paths did not function in the same way across all of the
parenting style groups.
To interpret the differences across the four parenting style
contexts, the significant chi-square for the each of the seven
significant paths described above were followed by six
pairwise analyses. In these analyses, the significant path
Table 4

was held invariant across two of the four parenting style


groups at a time (authoritative vs. authoritarian, authorita-
tive vs. indulgent, authoritative vs. uninvolved, authoritar-
66 MOUNTS

ian vs. indulgent, authoritarian vs. uninvolved, and indul- ation with adolescents’ Time 2 drug use for adolescents
gent vs. uninvolved). A significant chi-square on these from authoritative, authoritarian, and indulgent homes.
analyses would indicate that the path was different for those Parenting style group differences emerged for the path
two parenting style groups. Results of the pairwise analyses between adolescents’ Time 1 drug use and adolescents’
are presented using the subscripts in the second through fifth Time 2 drug use. Although for all groups, higher levels of
columns of Table 4. adolescents’ Time 1 drug use were associated with higher
The path between guiding and adolescents’ Time 1 drug levels of Time 2 drug use, this relation was much more
use was significantly different between adolescents who modest and nonsignificant for the adolescents from the
reported that their parents used the uninvolved style versus authoritarian parenting style group.
the other three parenting style groups. For adolescents who Finally, I found significant parenting style differences
reported that their parents used the uninvolved style, higher between friends’ Time 1 drug use and adolescents’ Time 2
levels of guiding were associated with higher levels of Time drug use. For adolescents from authoritarian families, hav-
1 drug use. High levels of guiding for adolescents in au- ing friends with higher levels of Time 1 drug use was
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

thoritative homes were associated with low levels of Time 1 related to higher levels of adolescents’ Time 2 drug use. The
This document is copyrighted by the American Psychological Association or one of its allied publishers.

drug use, whereas the authoritarian and indulgent groups results for the authoritative and indulgent groups were sta-
had a nonsignificant, but negative, association between tistically significant and similar in direction to the results for
guiding and adolescents’ Time 1 drug use. the authoritarian group. That is, having friends with higher
The path between monitoring and friends’ Time 1 drug levels of Time 1 drug use was related to higher levels of
use was significantly different for adolescents who reported Time 2 drug use. However, the magnitude of the relation
that their parents used an authoritative style versus the between friends’ Time 1 drug use and adolescents’ Time 2
authoritarian and uninvolved styles. For adolescents who drug use was significantly different from one another be-
reported that their parents used the authoritative style, tween adolescents from the authoritative and authoritarian
higher levels of monitoring were associated with having groups. The relation between friends’ Time 1 drug use and
friends with lower levels of Time 1 drug use. In contrast, adolescents’ Time 2 drug use was nonsignificant and was
monitoring was not associated with friends’ Time 1 drug also significantly different from the path associated with the
use characteristics for adolescents who reported that their authoritarian group.
parents used the authoritarian style, the indulgent style, or
the uninvolved style. Discussion
The path between neutrality and friends’ Time 1 drug use
The findings from the confirmatory factor analysis sug-
characteristics was significantly different between adoles-
gest that the four factors derived from earlier studies
cents who reported that their parents used an authoritarian
(Mounts, 2000, 2001) of the PMPI provided a good fit to the
parenting style and those who reported that their parents
data for this data set. Given that this measure is relatively
used the indulgent or uninvolved style. For adolescents who new, future studies should continue to explore the psycho-
reported that their parents used the indulgent style, higher metric properties of this measure for adolescents of different
levels of parental neutrality were associated with selecting ages and ethnic backgrounds. The analyses suggested that
friends with higher levels of Time 1 drug use. In contrast, the parenting style and the parenting practices constructs are
for adolescents who reported that their parents used the separate constructs. Future investigations of the linkages
authoritarian style, higher levels of neutrality were associ- between parenting style and parenting practices should as-
ated with selecting friends with lower levels of Time 1 drug sess whether the measures that are being used to assess
use. When parents used the authoritative or uninvolved these constructs are, in fact, distinct.
style, parent neutrality was not associated with friends’ The findings also support the notion that parental man-
Time 1 drug use. agement of peer relationships varies across the four parent-
The path between guiding and adolescents’ Time 2 drug ing style groups. Contrary to the hypotheses, there were no
use was significantly different between the uninvolved differences between the parenting style groups on prohibit-
group and the authoritative and indulgent groups. For ado- ing or guiding. Prohibiting occurred with similar frequency
lescents from uninvolved homes, high levels of guiding across all four parenting style groups, suggesting that this is
were associated with lower levels of Time 2 drug use, a strategy that all parents use, despite the fact that its effects
whereas guiding was not related to the Time 2 drug use for may vary within different parenting style contexts.
adolescents from authoritative and indulgent homes. The Analyses of the relations presented in Figure 1 suggest
paths between the adolescents from authoritarian homes and that four of the five parenting practices contributed to at
uninvolved homes were not significantly different from one least one of the outcome variables. Consistent with previous
another. literature, higher levels of monitoring were associated with
The path between prohibiting and adolescents’ Time 2 lower levels of Time 1 drug use. Surprisingly, higher levels
drug use was significantly different between the uninvolved of prohibiting were associated with higher levels of drug use
group and the remaining three groups. As can be seen in but only for Time 1. The directionality of this relation is
Table 4, higher levels of prohibiting were associated with unclear. It could be that when parents prohibit, they drive
higher levels of Time 2 drug use for adolescents in the their adolescents toward drug involvement. Alternatively, it
uninvolved group. Prohibiting had a nonsignificant associ- could be that parents are prohibiting in response to their
PARENTAL MANAGEMENT OF ADOLESCENT PEER RELATIONSHIPS 67

adolescents’ drug use. Higher levels of support were related contrast, the relation between monitoring and friends’ Time
to lower levels of Time 1 drug use. As expected, higher 1 drug use was not significant for any of the three remaining
levels of guiding were related to associating with friends parenting style groups. It could be that monitoring within
with lower levels of Time 1 drug use, whereas higher levels the context of the authoritative parenting style may be
of prohibiting were associated with having friends with interpreted differently by adolescents than would monitor-
higher levels of Time 1 drug use. Consistent with previous ing within the remaining three parenting style contexts. For
research, there were significant positive relations between example, monitoring may be viewed as being less intrusive
adolescents’ Time 1 drug use and friends’ Time 1 drug use. or, alternatively, because children of authoritative parents
In addition, as expected, adolescents’ Time 1 drug use and are more likely to internalize standards, they may be dem-
adolescents’ Time 2 drug use were significantly positively onstrating that this internalization has occurred within the
related. Further, friends’ Time 1 drug use was associated context of the peer relationship. These results in regard to
positively with adolescents’ Time 2 drug use, even when monitoring may also help explain some of the inconsisten-
controlling for the adolescents’ Time 1 drug use. This cies that have been found in the literature between moni-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

finding suggests that friends’ Time 1 drug use had an effect toring and peer relationships (Bogenschneider et al., 1998;
This document is copyrighted by the American Psychological Association or one of its allied publishers.

on adolescents’ drug use over time, as has been suggested in Fuligni & Eccles, 1993). At the very least, these results
previous research (Hawkins et al., 1992). suggest that monitoring may not function the same for all
The results provide modest support for the model pro- adolescents as it is usually assumed to operate.
posed by Darling and Steinberg (1993), in which parenting I found interesting differences across the groups for the
style moderates the relations between parenting practices path between neutrality and friends’ Time 1 drug use. In
and the outcome variable. Specifically, the analyses dem- particular, when adolescents reported that their parents used
onstrate that the model differs somewhat across the parent- the authoritarian style, higher levels of neutrality were as-
ing style contexts. I found significant parenting style group sociated with selecting friends with lower levels of Time 1
differences for the path between guiding and adolescents’ drug use. Given the high level of control that authoritarian
Time 1 drug use. The authoritative, authoritarian, and in- parents exert over their adolescents, this may override any
dulgent groups were significantly different from the unin- neutrality that parents may have in regard to peer relation-
volved group on this path. Higher levels of guiding were ships. That is, adolescents from authoritarian families may
associated with lower levels of Time 1 drug use for adoles- be so aware of the expectations of their parents, that even in
cents who reported that their parents used the authoritative, the absence of clear guidance in regard to peer relationships
authoritarian, or indulgent styles. In contrast, higher levels they still select friends whom they believe would be ap-
of guiding were associated with higher levels of Time 1 proved of by their parents. In contrast, for adolescents who
drug use for adolescents who reported that their parents reported that their parents use an indulgent or uninvolved
used the uninvolved style. It could be that the uninvolved style, higher levels of neutrality were associated with having
parents do not generally engage in guiding but resort to it friends with higher levels of Time 1 drug use. In this case,
when they discover that their children are involved in drug the overall tone of the parent– child relationships was very
use. Alternatively, the lack of parent– child relationships in permissive. In the absence of clear messages from the
the uninvolved group may have caused adolescents to do the parents about the acceptable range of peer relationships,
opposite of what their parents wanted them to do when their these adolescents selected friends with higher levels of drug
parents engaged in guiding. use.
The results are consistent with the work of Mize and The relation between guiding and adolescents’ Time 2
Pettit (1997), who found that parenting practices seemed to drug use was significantly different for the uninvolved
compensate for some of the weakness in parenting style. It group versus the authoritative and indulgent parenting style
is interesting that the results for the uninvolved group sug- groups. High levels of parental guiding were associated with
gest that the monitoring is not enough to compensate for the lower levels of Time 2 drug use for the adolescents in the
problems associated with the uninvolved style. Conversely, uninvolved group, whereas the relation between parental
when considering the adolescents who reported that their guiding and Time 2 drug use was nonsignificant for ado-
parents used the authoritative style, in all likelihood, mon- lescents in the authoritative and indulgent groups.
itoring becomes less important in determining drug use The results that examine the path between prohibiting and
because of the overarching effect of the parenting style on adolescents’ Time 2 drug use are similar to those for the
the adolescent. path between guiding and adolescents’ Time 1 drug use.
Monitoring functioned very differently within the author- Again, the relation between prohibiting and adolescents’
itative parenting style group versus the authoritarian, indul- Time 2 drug use was negative for the authoritative and
gent, and neglectful groups. Most of the existing research on authoritarian groups, whereas the relation was positive for
monitoring has not considered that monitoring may not the uninvolved group. Because these analyses included lon-
operate in a similar fashion for all parenting style contexts. gitudinal data, these results lend support to the explanation
The effects of monitoring on associating with drug-using that the low level of parental involvement created a climate
friends varied across the four parenting style groups. When whereby parents’ efforts to prohibit drove adolescents to-
adolescents reported that their parents used the authoritative ward higher levels of drug use at Time 2.
style, higher levels of monitoring were associated with There are several limitations of the present data set that
having friends with lower levels of Time 1 drug use. In also suggest directions for future investigations. First, the
68 MOUNTS

sample sizes for the four parenting style groups are some- who actively manage their adolescents’ peer relationships
what smaller than what is generally used when using struc- are increasing the likelihood that their children will associ-
tural equation modeling procedures, although analyses us- ate with peers who have lower levels of drug use (except for
ing samples of similar size have been conducted adolescents whose parents use an authoritarian style).
(Valentiner, Holahan, & Moos, 1994). The small sample Finally, the analysis of the model within the parenting
size may have increased the likelihood of a Type II error, style contexts suggested that certain parenting practices may
making it less likely that differences across the four parent- be taught to parents who may be labeled uninvolved. Bo-
ing style groups would be found. These analyses provided a genschneider and Stone (1997) suggested that even families
conservative test of the parenting style as moderator effect. at risk can learn more positive parenting practices. Some of
Second, the sample used in the present investigation con- the results from this investigation could be used to guide
tained only ninth-graders. Future investigations need to practitioners as they work with distressed families. In par-
expand the age range across adolescence. Developmental ticular, the adolescents from the uninvolved families tended
changes may affect the nature of management techniques to show a unique pattern of findings such that high levels of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

used, the effectiveness of the techniques in affecting peer guiding were associated with high levels of drug use at
This document is copyrighted by the American Psychological Association or one of its allied publishers.

relationships, and the effects of management practices on Time 1 but low levels of drug use at Time 2. The positive
the parental–adolescent dyad. Third, the samples in these relation between guiding and Time 1 drug use most likely
investigations were primarily composed of White adoles- occurred because parents were guiding in response to their
cents from lower-middle-class and middle-class back- children’s drug use rather than the guiding causing the drug
grounds. It is not clear the ways in which these management use. This explanation is further supported by the negative
techniques function across varying ethnic or socioeconomic relation between guiding and Time 2 drug use. This finding
status groups; however, some researchers are beginning to suggests that parents who have an uninvolved parenting
address this issue (Brown, Hamm, & Meyerson, 1996). style may benefit from information that encourages them to
Fourth, these investigations focused only on adolescents’ guide their children’s peer relationships. In addition, the
reports of parental management. Because parents and ado- results suggest that parents who have an uninvolved style
lescents may differ in their perceptions of the ways in which may be encouraging drug use by prohibiting their children
management occurs and how often it occurs, this is an from associating with particular adolescents. Therefore, pre-
important component of any research on direct parental vention and intervention efforts with uninvolved parents
influences on adolescent peer relationships. Fifth, Parke may include efforts to demonstrate to parents ways of being
(1992) suggested that research on parent–peer linkages involved in adolescents’ peers relationships without being
needs to consider bidirectional effects. Particularly as chil- prohibitive.
dren reach adolescence, peers may affect children, which
subsequently may affect parenting practices. More detailed
reports from parents and adolescents on specific instances of References
management and events just prior to management would Arbuckle, J. L., & Wothke, W. (1999). Amos user’s guide. Chi-
help illuminate the bidirectionality of effects. Finally, ado- cago, IL: SmallWaters Corporation.
lescents who are involved in antisocial activities often have Bauman, K., & Fisher, L. (1986). On the measurement of friend
older friends who have drawn them into antisocial activities behavior in research on friend influence and selection: Findings
(Magnusson et al., 1986). As a result, by restricting the from longitudinal studies of adolescent smoking and drinking.
sample to same-grade friends only, the peers who are the Journal of Youth and Adolescence, 15, 345–353.
most important to the adolescents, and likely to have the Bentler, P. M. (1990). Comparative fit indexes in structural mod-
most influence over adolescents, may not have been in- els. Psychological Bulletin, 107, 238 –246.
cluded in the sample. Future investigations may use a pro- Bentler, P. M., & Bonnett, D. G. (1980). Significance tests and
cedure in which adolescents and their very best friend from goodness of fit in the analysis of covariance structures. Psycho-
logical Bulletin, 88, 588 – 606.
any context participate in the research study. Berndt, T. (1979). Developmental changes in conformity to peers
and parents. Developmental Psychology, 15, 608 – 616.
Implications for Application and Public Policy Berndt, T. J. (1999). Friends’ influence on children’s adjustment to
school. In W. A. Collins & B. Laursen (Eds.), The Minnesota
The present investigation should be of interest to practi- Symposium on Child Psychology: Vol. 30. Relationships as de-
tioners for several reasons. First, it provides evidence that velopmental contexts (pp. 85–107). Mahwah, NJ: Erlbaum.
parents do have an impact on their children’s involvement Billy, J., & Udry, J. (1985). Patterns of adolescents’ friendship and
in drug use, even when accounting for the effects of friends. effects on sexual behavior. Social Psychology Quarterly, 48,
This is good news for parents. Although children may not 27– 41.
always select friends who are engaging in desirable activi- Bogenschneider, K., & Stone, M. (1997). Delivering parent edu-
cation to low- and high-risk parents of adolescents via age-paced
ties, specific parenting practices may prevent adolescents newsletters. Family Relations, 46, 123–134.
from being unduly influenced by these friends. Bogenschneider, K., Wu, M., Raffaelli, M., & Tsay, J. (1998).
Second, this investigation provides evidence that parents Parent influences on adolescent peer orientation and substance
can have an effect on the types of friends that adolescents use: The interface of parenting practices and values. Child De-
are associating with. Given that there is a strong relation velopment, 69, 1672–1688.
between peer behavior and adolescent behavior, parents Bronfenbrenner, U. (1979). The ecology of human development:
PARENTAL MANAGEMENT OF ADOLESCENT PEER RELATIONSHIPS 69

Experiments by nature and design. Cambridge, MA: Harvard management of peer relationships. Journal of Early Adolescence,
University Press. 21, 92–122.
Brown, B. B., Hamm, J. V., & Meyerson, P. (1996, March). Mounts, N. S., & Steinberg, L. (1995). An ecological analysis of
Encouragement, empowerment, detachment: Ethnic differences peer influence on adolescent grade point average and drug use.
in approaches to parental involvement with peer relationships. Developmental Psychology, 31, 915–922.
Paper presented at the biennial meeting of the Society for Re- Parke, R. D. (1992) Epilogue: Remaining issues and future trends
search on Adolescence, Boston, MA. in the study of family–peer relationships. In R. D. Parke & G.
Brown, B. B., Mounts, N. S., Lamborn, S. D., & Steinberg, L. D. Ladd (Eds.), Family–peer relationships: Modes of linkage (pp.
(1993). Parenting practices and peer group affiliation in adoles- 425– 438). Hillsdale, NJ: Erlbaum.
cence. Child Development, 64, 467– 482. Parke, R. D., & Ladd, G. (Eds.). (1992). Family–peer relation-
Burger, G. K., & Armentrout, J. A. (1971). A factor analysis of ships: Modes of linkage. Hillsdale, NJ: Erlbaum.
fifth and sixth graders’ reports of child-rearing behavior. Devel- Parke, R. D., & O’Neil, R. (1999). Social relationships across
opmental Psychology, 4, 483. contexts: Family–peer linkages. In W. A. Collins & B. Laursen
Darling, N., & Steinberg, L. (1993). Parenting style as context: An (Eds.), The Minnesota Symposium on Child Psychology: Vol. 30.
integrative model. Psychological Bulletin, 113, 487– 496. Relationships as developmental contexts (pp. 211–239). Mah-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Dornbusch, S. M., Carlsmith, J. M., Bushwall, P. L., Ritter, P. L., wah, NJ: Erlbaum.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Leiderman, P. H., Hastorf, A. H., & Gross, R. T. (1985). Single Parker, C. P., Baltes, B., & Christiansen, N. D. (1997). Support for
parents, extended households, and the control of adolescents. affirmative action, justice perceptions, and work attitudes: A
Child Development, 56, 326 –341. study of gender and racial– ethnic group differences. Journal of
Fuligni, A. J., & Eccles, J. S. (1993). Perceived parent– child Applied Psychology, 82, 376 –389.
relationships and early adolescents’ orientation toward peers. Savin-Williams, R., & Berndt, T. J. (1990). Friendship and peer
Developmental Psychology, 29, 622– 632. relations. In S. S. Feldman & G. R. Elliott (Eds.), At the thresh-
Glasgow, K. L., Dornbusch, S. M., Troyer, L., Steinberg, L., & old: The developing adolescent (pp. 277–307). Cambridge, MA:
Ritter, P. (1997). Parenting styles, adolescents’ attributions, and Harvard University Press.
educational outcomes in nine heterogeneous high schools. Child Schaefer, E. S. (1965). Children’s reports of parental behavior: An
Development, 68, 507–529. inventory. Child Development, 36, 413– 424.
Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and Stattin, H., & Kerr, M. (2000). Parental monitoring: A reinterpre-
protective factors for alcohol and other drug problems in adoles- tation. Child Development, 71, 1072–1085.
cence and early adulthood: Implications for substance abuse Steinberg, L. (1986). Latchkey children and susceptibility to peer
prevention. Psychological Bulletin, 112, 64 –105. pressure: An ecological analysis. Developmental Psychology, 22,
Kandel, D. (1978). Homophily, selection, and socialization in 433– 439.
adolescent friendships. American Journal of Sociology, 84, 427– Steinberg, L., Lamborn, S., Darling, N., Mounts, N., & Dornbusch,
436. S. (1994). Over-time changes in adjustment and competence
Ladd, G., & Golter, B. (1988). Parents’ management of preschool- among adolescents from authoritative, authoritarian, indulgent,
er’s peer relations: Is it related to children’s social competence? and neglectful families. Child Development, 65, 754 –770.
Developmental Psychology, 24, 109 –117. Steinberg, L., Lamborn, S., Dornbusch, S., & Darling, N. (1992).
Lamborn, S., Mounts, N. S., Steinberg, L., & Dornbusch, S. Impact of parenting practices on adolescent achievement: Au-
(1991). Patterns of competence and adjustment among adoles- thoritative parenting, school involvement, and encouragement to
cents from authoritative, authoritarian, indulgent, and neglectful succeed. Child Development, 63, 1266 –1281.
families. Child Development, 62, 1049 –1065. Steinberg, L., & Silverberg, S. (1986). The vicissitudes of auton-
Magnusson, D., Stattin, H., & Allen, V. (1986) Differential mat- omy in early adolescence. Child Development, 57, 841– 851.
uration among girls and its relation to social adjustment: A Urberg, K. A., Cheng, C., & Shyu, S. (1991). Grade changes in
longitudinal perspective. In P. Baltes, D. Featherman, & R. peer influence on adolescent cigarette smoking: A comparison of
Lerner (Eds.), Life-span development and behavior (Vol. 7, pp. two measures. Addictive Behaviors, 16, 21–28.
135–172). Hillsdale, NJ: Erlbaum. Valentiner, D. P., Holahan, C. J., & Moos, R. H. (1994). Social
McCord, J. (1990). Problem behaviors. In S. Feldman & G. Elliot support, appraisals of event controllability, and coping: An inte-
(Eds.), At the threshold: The developing adolescent (pp. 414 – grative model. Journal of Personality and Social Psychology, 66,
430). Cambridge, MA: Harvard University Press. 1094 –1102.
Mize, J., & Pettit, G. S. (1997). Mothers’ social coaching, mother– Vernberg, E. M., Beery, S. H., Ewell, K., & Abwender, D. A.
child relationship style, and children’s peer competence: Is the (1993). Parents’ use of friendship facilitation strategies and the
medium the message? Child Development, 68, 312–332. formation of friendships in early adolescence: A prospective
Mize, J., Russell, A., & Pettit, G. S. (1998). Further explorations of study. Journal of Family Psychology, 7, 356 –369.
family–peer connections: The role of parenting practices and Youniss, J., DeSantis, J., & Henderson, S. (1992). Parents’ ap-
parenting style in children’s development of social competence. proaches to adolescents in alcohol, friendship, and school situa-
In P. T. Slee & K. Rigsby (Eds.), Children’s peer relations (pp. tions. In I. Sigel, A. V. McGillicuddy-Delisi, & J. J. Goodnow
31– 44). New York: Routledge. (Eds.), Parental belief systems: The psychological consequences
Mounts, N. S. (2000). Parental management of adolescent peer for children (pp. 199 –216). Hillsdale, NJ: Erlbaum.
relationships: What are its effects on friend selection? In K.
Kerns, J. Contreras, & A. Neal-Barnett (Eds.), Family and peers: Received June 20, 2000
Linking two social worlds (pp. 167–193). Westport, CT: Praeger. Revision received January 26, 2001
Mounts, N. S. (2001). Young adolescents’ perceptions of parental Accepted August 3, 2001 䡲

You might also like