1
Susceptibility to Peer Influence
Leaders and Followers in Adolescent Peer Relationships:
Susceptibility to Peer Influence as a Predictor of
Peer Pressure, Risky Behavior and Depression
Joe Allen
Christy McFarland
University of Virginia
Paper presented at Biennial Meetings of the Society for Research in Child Development, Tampa, FL April 2003
(Talk is keyed to slides on accompanying powerpoint presentation.)
I.SLIDE#1)Title:
II. Introduction
SLIDE#2) : Autonomy : There is an enduring paradox that characterizes the study of social development in adolescence.
1.On the one hand, forming strong, lasting peer relationships is critical to normal social development.
2. Yet, adolescents’ conformity to the negative norms of the peer groups that are becoming increasingly important also appears as one of the major risk factors linked to negative outcomes ranging from delinquency and substance abuse to risky sexual behavior.
Said Differently, we know that all adolescents are exposed to at least some deviant behavior from peers.
a.Yet, we also know that exposure to peers engaging in deviant behavior isn’t the same thing as being influenced by those peers.
B. What we don’t yet know, and what may be critical to resolving this paradox, is how to understand which adolescents are most likely to join in vs. avoid the deviant behavior present to some degree in almost all peer groups.
1.Negative peer pressure is clearly a huge concern in adolescence--and yet it doesn’t affect all kids equally.
2.If we can understand the processes that lead some kids to be negatively influenced by their peers, whereas others aren’t perhaps we can begin to think about ways of addressing problems of negative peer influences.
3.So, how do we think about this?
4. We think that adolescents’ handling of peer influences can be very productively conceptualized as a reflection of their general progress in the developmental task of establishing autonomy in social interactions.
C.In this respect we’d like to suggest a partial answer to what we see as a fundamental question: SLIDE#3)Overarching Question :
1.The question being: What does it take for an adolescent to be “competent” in handling interactions with his/her broader peer group?
2.And among the many different areas of competence that could be considered, we’d like today to suggest that the role of autonomy processes in peer interactions is one critical aspect of social competence with peers that hasn’t yet received sufficient attention.
D.Previously, we’ve studied an adolescent’s ability to develop verbal autonomy while maintaining positive relationships WITHIN THEIR FAMILIES as a critical task of social development.
1.The study that I’ll talk about today, examined the idea that adolescents’ handling of autonomy issues with peers may be an equally critical task in two respects:
a.First, in allowing adolescents to resist the range of problem behaviors associated with negative peer influences in adolescence.
b.And second,k we’d like to present data to suggest that a lack of autonomy with peers, as we’ll measure it, isn’t simply a risk factor for problematic behavior, but may actually have significantly broader implications for a range of aspects of psychosocial development in adolescence.
SLIDE#4) : Turning to measurement issues: in spite of the potential importance of understanding susceptibility to peer pressure (or lack of autonomy with peers) in adolescence,
2.with the exception of one early study of peer conformity virtually
E.noresearch has been conducted on susceptibility to peer influence other than via adolescent self-report , even though self reports are least likely to be reliable when assessing highly charged social interactions of this type.
1. Further, even the extant self-report measures of peer pressure have either suffered fundamental psychometric problems, or have confounded the assessment of adolescents’ susceptibility to peer pressure with their adolescents’ willingness to engage in deviant acts as they grow older.
a.that is they’ve assessed peer pressure, but always and almost only in the context of would an adolescent be willing to engage in deviant behaviors if pushed by a peer, so that there’s been no way to disentangle increasing acceptance of deviant behavior from susceptibility to peer pressure.
F.In this study, we sought to develop an experimental task to assess the degree to which peers could be influenced by a close friend in a neutral context (i.e. one where deviant behavior was not a part of th task)
We then sought to assess the relationship of susceptibility to peer influence in this neutral context to measures ranging from deviant behaviors to broader markers of psychosocial functioning.
III.Methods
A.SLIDE#5)Sample We assessed a community sample of 168 adolescents, drawn from the public school system of Charlottesville, Virginia,along with their parents and friends,
1.evenly divided between males and females,
2.initially assessed at age13, and re-assessed annually over the next two years
3. The sample was representative of the population of the surrounding community in both socioeconomic and racial/ethnic terms as you can see on the overhead.
We interviewed and observed adolescents
a.alone,
b.with their parents,
c.and with their best friends,
d.and we also asked adolescents to name additional members of their peer group who we invited in and also interviewed.
In seeking to devise a measure of susceptibility to peer influence we had several goals:SLIDE#6)Goals a :
4.clickFirst, We wanted to develop a measure that would tell us something about social development broadly defined--and not just about problematic behavior.
5.clickIn a similar vein, we did not want our assessment to build in a confound with whether or not an adolescent was willing to engage in a deviant behavior (i.e. we were looking for a neutral task, in other words).
a.If our theoretical perspective is on target, then we expect that lacking autonomy in relation to one’s peers is a general phenomenon, and not something that is just specific to problematic behavior--
b.even though that’s where most assessments to date have been focused.
6.clickFinally, and obviously, we wanted to move beyond reliance upon adolescent self-reports.
To do this we devised a task about which teens could disagree even with their closest friends.SLIDE#7) Msrs: Susceptibility:
7.The experimental task is a hypothetical “sinking ship” type of dilemma. We called it the “Mars Task”. In this case, teens are on a trip back from mars when an accident leaves them with too few spacecraft to rescue all the passengers on board.
8.They listen to a dramatic reading of the situation on an audiotape, then separately they make decisions about which of the hypothetical characters on board should get highest priority for rescue.
9.clickAfter they’ve done this individually, they come together with their best friend, who has been completing the same task separately, and are told to try to come to agreement about who should be rescued.
10.Although this task is hypothetical, in many ways, it taps into exactly the type of hypothetical discussions that young adolescents have every day--
a.for example, if you had to give up your eyes or your ears which would it be
i.That one’s from an extensive qualitiative analysis of conversations overheard in our car pool…and I can only hope its hypothetical.
or a favorite from my adolescence, if a wrestler and a boxer fought in a ring, who would win?
In short, these discussions are hypothetical, but they are also one of the ways that adolescents learn to grapple with their social worlds.
11.clickwe then simply calculated the percentage of the time the adolescents’ position “lost out” in the final consensus to their friend’s position, to come up with our measure of susceptibility to peer influence.
12. And, as we might expect, teens and peers each changed the other’s mind about half the time.
13.4:00
IV.Results
A.So, we obtained this brief experimental assessment--the question is: Is it meaningful in terms of larger patterns of social behavior?
B.First, we looked to see how our experimental assessment matched up with close peer reports of actual susceptibility to negative peer influences
1.SLIDE#8) : We assessed actual negative peer influences with a measure that we developed for this project in which we asked, not the teen, but their closest friend, to tell us about the extent to which the target teen was actually influenced by their best friend in a variety of negative ways, as outlined on the slide. ---…..
2.SLIDE#9) : We found that our experimental assessment was indeed related, albeit somewhat weakly to close peers’ reports of influencing our target teens’ to engage in negative behavior.
a.[I should note that in each model we covary out any demographic factors that were related to either predictors or outcomes in that model.]
3.We see these findings both provide some construct validity for our new measure, but also suggests that susceptibility in a neutral context may share something in common with susceptibility in the more critical context of actual behaviors.
C.Turning next to Look at some actual behaviors of our teens, we begin with alcohol and drug use
1.SLIDE#10) :
2.We asked adolescents about problems they had had with alcohol and drug use over the past year, using a 4-item scale we developed using Susan Harter’s format from the Self Perception Profile to reduce pulls for social desirability. A sample item, as you can see is on the slide.
D.SLIDE#11) : We see that after accounting for gender and race of our adolescents
E.susceptibility to peer influence was significantly related to teens’ difficulties with alcohol and drug use.
1. Thus, teens who could be more easily persuaded to change their minds in a brief hypothetical discussion also were more likely to have difficulties with alcohol and drug use outside of the lab.
F.Now, one of the obvious possible mechanisms to explain this link is that these teens were actually influenced by their peers’ behavior to engage in more alcohol and drug use.
To examine this possibility, we also asked 2 peers of the adolescent (their best friend and an additional person within their circle of closest friends) to tell us about their own alcohol and drug use, SLIDE#12)Msrs-Peer Use :
1.In this case, simply about whether or not these peers had previously used alcohol or marijuana
2.Thus we have a yes/no question for each of alcohol and marijuana for each of 2 friends for a 0-4 point scale for PEERS’ use of alcohol and Marijuana
SLIDE#13) : When we add these into our prior model, we see first, that peers’ alcohol and drug use does predict adolescents’ levels of use
3.(this isn’t surprising, although this is one of a relatively small # of studies that examines this association without using the teen to reports of both their own and others’ drug use).
SLIDE#14) : More strikingly though, we find an interaction between susceptibility to peer influence and peer drug use in predicting adolescents difficulties.
SLIDE#15) : As the figure depicts, for teens who were low in susceptibility, it didn’t much matter what their peers’ levels of use were. But for teens who were high in susceptibility, there was a strong direct correlation between their peers’ use of alcohol and substances and their own use.
4.So our teens who were followers in the “Mars task” did in fact appear to be following in their peers’ footsteps, whereas for our teens who were leaders, there was no relation between their peers’ use and their own use.
V.Now without going into too much detail, let me note that susceptibility was also related to other areas of problematic behavior (although we have not found the same interactional finding with these other behaviors as we did with substance use).
A.But we do find that susceptibility predicts higher levels of general externalizing behaviors SLIDE#16) :
1.now we use maternal report of externalizing behaviors on the child behavior checklist, which has the advantage of giving us an assessment that not only isn’t the adolescent’s, but that is totally separate from the peer context.
2.SLIDE#17) : And that susceptibility to peer influence is linked to greater likelihood of becoming sexually active at an early age (these were 13 and 14 year olds we were assessing).
a.Notably, using logistic regression, our teens who were 1 s.d. above the mean in susceptibility had twice the likelihood of becoming sexually active by age 14 as our teens who were average in susceptibility.
B.So, we now have evidence that a general susceptibility to peer influence is indeed linked to a variety of specific problem behaviors,
C.and we have some evidence that more susceptible adolescents are particularly vulnerable when they are with peers whose behavior is more problematic
VI.What we’d like to do now is turn to looking at a Larger hypothesis:
A.which is that susceptibility to peer influence--or lack of autonomy with peers, as we’re thinking of it--actually reflects significantly broader difficulties in psychosocial functioning.
B.We have two types of data we look at here.
1. One has to do with the quality of the best friendship of the susceptible teen,
2.and the other has to do with levels of depression.
C. We assess close friendship quality using the close friendship scale of Harter’s Self Perception Profile as shown on the slide SLIDE#18) : and we combine both teen reports about themselves and peers’ reports about our target teen to get our overall scale.
SLIDE#19) : We find that susceptibility to peer influence is also linked to lower quality close friendships.
D.We then looked at the stability of the friendship by asking whether the teen chose to, or was able to re-recruit the same close friend to return to our lab 1 year later for a follow-up visit.
1.SLIDE#20) : And again, we find that susceptible teens were less likely to recruit the same friend to come back into our labs with them the following year.
2.To interpret the results of this analysis,
a.for the teen who was 1 s.d above the mean in susceptibility to peer influence, they were only 2/3’s as likely to get their same friend to return the following year as was the average teen.
3.Most straightforwardly, these results show that easily influenced teens have lower quality and less stable friendships.
E.Now, we’ve often argued that autonomy and relatedness in adolescence not only aren’t opposing qualities of relationships but rather tend to go hand in hand.
F.We see these data as supporting this perspective, and while we don’t know which way the causal arrow goes--
1.whether its that teens who are insecure about their friendships are less willing to push for their viewpoints, or that teens who are overly susceptible to peer influence have friends who take them less seriously--
G.in either case, its clear that lack of autonomy and lack of strong relationships are closely interconnected in the peer world.
VII.Finally, as our last piece of evidence that susceptibility to peer influence may do more than “just” highlight a risk of negative behaviors,
1.We examined the Childhood Depression Inventory scores for our adolescents. Here we found that while at Time 1 (age 13), depressive symptoms were not related to susceptibility,
2.SLIDE#21) : by age 14 as you’ll see on the slide, susceptibility did predict levels of depressive symptoms, and in fact predicted changing levels of depressive symptoms over time SLIDE#22) :
B.It may be that insecurity about one’s friendships, and perhaps even some passivity that comes with low-grade depressive symptoms might leave the adolescent feeling as though they are failing to keep up with a fundamental developmental task.
1.Alternatively, establishing relationships in which one cannot assert one’s own influence effectively has been identified within interpersonal process theories of depression as a central aspect of depression in adolescence and adulthood.
C.In either case, it is clear that susceptibility to peer influence at least marks a vulnerability to increasing levels of depressive symptoms over time in early adolescence.
Discussion SLIDE#23) : So to summarize:
D.This study developed an experimental approach to assessing early adolescents’ susceptibility to peer influence and found it to be linked to actual negative influences upon the adolescents, to their difficulties with substance use and their levels of externalizing problem behaviors
E. And in a broader sense, links were found not just to problematic behavior, but ,also to difficulties in relationships with close friends and to future depressive symptoms.
F.The relations observed were obtained across a variety of methods and raters.
SLIDE#24)SUMMARY CONT : It is also noteworthy that our assessment of susceptibility is to some extent symmetrical so that while high scores indicated high susceptibility to peer influence, low scores reflected high levels of influence exerted upon one’s close friend (in other words: leadership).
1.This means that adolescents who were more successful in asserting their influence with their close friends displayed lower levels of the problematic behavior and adaptational difficulties.
VIII.Limitations
A.Clearly, these data are correlational in nature, and even though they are longitudinal in places, they obviously cannot support causal claims.
B.Also, I should also note that our measure was designed in such a way that the result is influenced both by the teen and by their close peer (its capturing a property of the dyad)…so in that sense, only some of the variance of this measure is from the teen, and while we have some reason to think that the teen may determine more of the variance in this measure as we use it, nevertheless these correlations thus all will tend to some extent to underestimate the true effects of susceptibility, because our measure is not “purely” determined by the adolescent.