How Does Sex Fit Into a Broader Context

1

Observed Autonomy and Connection with Parents and Peers as Predictors of Early Adolescent Sexual Adaptation

Joseph P. Allen

Felicia Hall

University of Virginia

Talk Presented at the Biennial Meetings of the Society for Research in Adolescence, New Orleans, Louisiana, April 11, 2002

SLIDE#1)Title: I’d like to first begin today by acknowledging my co-author Felicia Hall, and the numerous other current and former students listed on the slide who contributed to this paper.

A.In keeping with the theme of this symposium, our approach to understanding adolescents’ sexual and dating behaviors is to view them as reflecting social skills and competencies that adolescents are developing OUTSIDE of romantic relationships—

B.SLIDE#2)Theoretical Perspective We begin with the theoretical perspective that adolescents ideally are making a gradual transition from getting their needs for support, affection and intimacy met by parents… to getting these needs met in same-gender friendships, and eventually in romantic relationships over time—and that this transition is one of the most challenging and potentially stressful tasks of early- to mid-adolescence

1.This period…from around 7th through 11th grades, is one in which adolescents are actively learning NOT to rely as heavily on parents for physical affection, and for close intimate interaction.

2.YET, they have also NOT yet developed strong peer outlets to replace what they are foregoing within their families.

3.From this perspective, we think adolescents may face what we’ve termed an “intimacy gap”.

a.Younger children can turn to parents for extensive amounts of affection, intimacy and support.

b.Older adults develop strong friendships and romantic relationships to meet their needs for intimacy.

c.Early to mid-adolescents, however, don’t necessarily have any easy place to turn to meet needs for emotional intimacy, physical affection and support.

d.Parents may fill these needs to a degree, and same-gender peers increasingly have a role to play, but…

e.At an unusually vulnerable time, early adolescents have fewer sources of close emotional support than most individuals do at almost any other period in life.

4.We see this as particularly relevant to understanding dating and sexual behavior during this period, because adolescents who are most lacking in support from parents and same-gender peers seem most likely to turn to romantic relationships for such support--

5.---even though in early adolescence, romantic and sexual relationships are particularly poorly suited to provide this support (both as a result of their instability and due to the risks that come with early sexual activity).

C.Based on this perspective, we hypothesized that early adolescents who were having the greatest difficulty establishing intimate and supportive relationships first with parents and then with close same-gender friends, would be most likely to turn to precocious dating and sexual activity—to address this intimacy gap.

1.SLIDE#3) Central QuestionOur primary question then was:

a.Do high quality close relationships with parents and peers predict a slower onset of dating and sexual behavior in early adolescence?

b.And if so, which aspects of these relationships are most important?

II.Methods

A.SLIDE#4)Sample To address these questions, we assessed a community sample of adolescents drawn from the public school system of Charlottesville, Virginia.

1.The sample consisted of 148 adolescents, evenly divided between males and females, initially assessed at age13 on average, and re-assessed annually over the next two years.

2.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.

B.SLIDE#5)Measures-Dating Our measure of dating behavior consisted of 5 questions regarding the extent of adolescents’ early dating experience).

1.Adolescents described both dating and sexual behaviors each year, in a computer-assisted interviewing format that maximized adolescents’ anonymity in providing information.

2.Basic descriptive statistics for our dating measure at ages 13, 14 and 15 are also shown on the slide and reflect in essence that adolescents almost all report having had a boyfriend/girlfriend and some early romantic activity (e.g., hand holding) by age 13, but are now increasing in #’s and frequencies of romantic activities over the following years. I’ll say more about the frequencies of these specific behaviors in just a minute.

C..SLIDE#6)Measures-Sex Although we have a number of detailed questions we’re using to track adolescents’ sexual activity, at this age, given the low base rates of sexual activity, we focus simply on the presence or absence of a history of consensual sexual intercourse as our primary marker of sexual activity.

a.You can see these base rates at the bottom of the slide, and their steady increase over time. We’ll focus today on those adolescents who had become sexually active at or before age 14 as our group of primary interest, in part because our age 15 data is still not quite complete.

D.Given that the onset of actual sexual behavior was relevant only to a small portion of our sample, we also wanted to get at relevant attitudes toward sexual behavior. SLIDE#7)Attitudes

1.We specifically examined adolescents’ (and parents’) opinions about the age at which it is OK to begin various kinds of romantic behaviors.

a.We assessed this for both mothers and adolescents, by presenting them with a series of behaviors, ranging from having a boyfriend/girlfriend, to going out on dates, to having sex, and by asking the age at which they considered it OK, for a youth to engage in these behaviors.

b.We average their responses into a measure of overall liberal attitudes (e.g., extent to which they see these behaviors as being “OK” at earlier ages).

c.When we take a preliminary look at these attitudes toward romantic behavior, On the one hand, we see mothers’ and teens’ attitudes being significantly correlated with each other r=.52, suggesting a great deal of correspondence between maternal and teen values in this regard. But, as you’ll see on the next slide…

2.SLIDE#8)maternal vs. teen attitudes there were actually very significant mean differences in the ages at which teens’ and their parents thought various behaviors were acceptable.

a.Teens’ responses were correlated with their mothers’ responses, but teens’ consistently saw behaviors as acceptable about 2 years earlier than did their parents.

3.SLIDE#9)maternal attitudes vs. teen behaviors : And if we turn to the next slide, looking at the actual incidence of these behaviors in our sample, we see that teens’ behaviors more match up with their own attitudes far more than they match up with their parents’ attitudes.

a.94% of the teens in our samples have had someone that they referred to as a boyfriend/girlfriend by age 13,

b.57% have been out on dates (we didn’t’ ask about alone vs. in groups)

c.64% have kissed a girl or boy.

4.SLIDE#10)Attitudes  dating Do these attitudes actually matter in terms of teens’ behavior? On this next slide, we see that this measure of teens’ attitudes demonstrates some real evidence of predictive validity, in that attitudes at age 13 in fact predict changes in dating behavior over the following 2 years.

a.suggesting it may well be a valid measure to use to assess attitudes in a sample of early adolescents, where critical sexual and dating behaviors are just starting to become relevant.

5.So, in sum, we rely on 3 measures of romantic activity…dating behavior, initiation of sexual intercourse, and a measure of teens’ liberal vs. conservative attitudes toward romantic and sexual behaviors.

E.We also examined two major classes of social behavior with parents and peers as potential predictors of adolescents’ sexual and romantic attitudes and behaviors. SLIDE#11)Predictors

F.First, we looked at autonomy –related behaviors by observing how parents and their teens handled a disagreement. SLIDE#12)Autonomy

1.We did this by using the Autonomy & Relatedness Coding system, which captures the extent to which a dyad—mother and teen in this case--can handle a disagreement while still preserving each member’s autonomy and sense of relatedness.

G.SLIDE#13)Supp. Beh. Task:We also looked at the “connection” aspect of relationships by examining both peer and parenting Support and Connection using a Supportive Behavior Task in which teens are told to discuss a topic with their parents (or peers) about which they would like to get some help.

H.We code these videotaped discussions for a number of different qualities, as outlined on the overhead, using a coding system adapted from work by Crowell & Waters & by Markman, Julien, & Colleagues SLIDE#14)Supp. Beh. Coding

I.These qualities include:

1.Emotional support given by the parent or peer.

2.The degree of each party’s engagement in the interaction

3.The listener’s (whether parent or peer’s) ability to successfully interpret what the teen is requesting in the interaction.

4.And finally, the Teens’ Satisfaction with the interaction

III.Turning, now to our results, we first look at Autonomy Tasks in the Parent-teen relationship as a predictor of dating and sexual behavior, we get a strikingly clear set of findings as you’ll see on the next slide. SLIDE#15)Autonomy non-findings :

A.In essence, we find no predictions from autonomy in interactions with parents to adolescents’ sexual or dating behaviors or attitudes.

B.To be clear, we’ve found that these measures of autonomy processes predict numerous other aspects of social development during adolescence, from delinquency to depression and social withdrawal. They do not, however, appear linked to dating and sexual behavior.

IV.Turning to our Measures of CONNECTION, that is our observation of parent and peer interactions in a supportive behavior task, we see a much clearer pattern of findings begin to emerge.

V.SLIDE#16) Dating & Satisfaction with mother Our first set of findings is with respect to DATING & Mother-teen interactions.

A.We see that early teen dating behavior is linked to adolescents’ lack of satisfaction with their mother’s efforts to be supportive in a supportive interaction task.

1.In essence, teens are asking their mothers for help in the task, but aren’t satisfied with the help they are getting.

2.And it is these teens who are most likely to begin dating earlier and in a more intensive way by age 14.

B.We do not, however, see any predictions of dating behavior from interactions with peers, nor from any other aspects of interactions with mothers.

VI.When we turn from DATING TO SEXUAL BEHAVIOR, we see a different but somewhat consistent picture.

A.Here, the best friend relationship seems to be the most consistent and robust predictor of later sexual behavior.

B.SLIDE#17)Peer Engagement First, we see two predictions obtained simply by observing the behavior, not of our target teen, but of their same gender close friend who participated in these interactions—

C.In essence we’re asking if we can predict teens’ likelihood of becoming sexually active by observing what they evoke from their same gender best friend in an interaction task.

1.First, the extent to which that peer became highly engaged with our target teen during the interaction predicted a lower likelihood of becoming sexually active by age 14

2.If you’re best friend is highly engaged with you in early adolescence—you are less likely to become sexually active at an early age.

D.SLIDE#18) : Second, we find that one’s close friends’ skill at interpreting the problem the adolescent brings up in the interaction is also predictive of lower levels of sexual experience by age 14.

1.I should mention that we’re using logistic regressions in these analyses, and while we present standardized B weights for ease of interpretation of effects for those who are used to this metric, the odds ratios are probably more informative.

2.These in essence say that teens whose peers are high in interpretation skill…in this case about 1 s.d. above the mean, have about half the likelihood of having sex by age 14 as those whose peers are merely average at interpreting their friends’ needs in a supportive behavior task.

3.Summarizing these first two findings, in some sense, we find its not who you are, its who your friends are (or more precisely, what they are like when they interact with you) that best predicts the onset of teen sexual activity.

4.Early adolescents who have same-gender friends who listen and understand and are engaged in interactions are less likely to seek out early sexual experiences.

E.When we looked at our target teens’ own behaviors in these interaction tasks, we see several complementary predictions SLIDE#19)Engaged Interaction (teens))

1.Teens who themselves were more engaged with their peers were also less likely to be sexually active.

F.SLIDE#20)Satisfaction Similarly, and also parallel to our peer measures, teens who were more satisfied with interactions with their peers were also less likely to be sexually active.

1.We tried adding these variables simultaneously into analyses—to see if they perhaps contributed unique variance to explaining sexual behavior outcomes

a.Unfortunately, our sample size of sexually active kids was still too small to provide much power for these analyses at age 14, and so we don’t find any reliable, additive effects.

G.In sum, though, what we’ve got is a clear finding, with a number of different indices, showing that teens who are able to establish relationships with their same-gender best friends that are of a high emotional quality, do not seek out early sexual experiences. Conversely, adolescents who do not or cannot seek out and obtain these relationships are most likely to engage in early sexual behavior.

1.We tested for moderating effects of gender in these analyses, but none were found—which we think is interesting in that it suggests that the need for emotional engagement, and its link to sexual behavior, appears equally powerful and applicable to males as it does to females in our sample.

H.Looking now at our teen-parent interactions, we find one very similar prediction from mother-adolescent interactions to sexual behavior SLIDE#21)Maternal Engagement

1.Teens who have mothers who are more engaged in interactions with them had lower levels of sexual behavior.

2.We think the best way to think about this slide is to put these findings together with our data on the relation of maternal engagement in the mother-teen task, and teen-engagement in the teen-peer task.

3.SLIDE#22) : What you’ll see on this next slide is our attempt to do just that.

4.This slide shows one possible way of viewing these data which is to say that mothers’ engagement with their teens is related to teens’ engagement with their peers, which in turn predicts a lower likelihood of being sexually active by age 14.

5.There are other interpretations of these data that are possible of course, but this view suggests that perhaps mothers’ have a role in teens’ sexual behavior at this age, but its primarily in setting the stage for the type and quality of peer relationships the adolescent will be able to establish.

I.Finally, we’d like to turn to our predictions of teens’ attitudes about the age at which various behaviors are OK to begin. We used these, you’ll recall, as proxies for sexual behavior given that most of our sample wasn’t becoming sexually active during the first two years of our study.

1.In general, we find that largely the same predictors that we showed with respect to teen sexual behavior also apply to teen attitudes toward romantic behavior and in the same directions.

2.The one thing our attitude measure lets us do, however, is use its more friendly psychometric properties to do things like look at change over time, and this helps us detect some slightly more subtle variations on the findings we’ve just reported.

J.For example, we found that mothers’ engagement with their teens interacted with mothers’ own values and attitudes in predicting teens’ attitudes toward romantic behaviors SLIDE#23) predicting teen attitudes:

1.Specifically, what we see is that maternal engagement is associated with teens having more conservative attitudes about romantic behavior by age 15.

a.This is completely consistent with what we’ve been reporting above, but we also find that maternal attitudes interact with maternal engagement to predict teens’ attitudes, as illustrated on the next slide. SLIDE#24) graph

2.Specifically, what we see on the graph is that lack of engagement particularly matters for mothers’ with very liberal attitudes toward teen sex…

3.In these cases, with disengaged mothers who hold liberal attitudes, teens’ have the most liberal attitudes toward sexual and romantic behaviors of any teens in our sample

a.I should also note that our preliminary data on actual sexual activity, as opposed to attitudes, yields exactly the same pattern of results.

4.15:40

K.In our analyses of predictors of teen attitudes, we also find that we can extend our prior findings on the role of adolescents’ engagement with peers. SLIDE#25) :

1.By looking now at adolescents’ attitudes, we find that engagement with peers actually predicts CHANGE in these attitudes over time.

a.That is, the more engaged a teen is with their close friend in getting support at age 13, the more likely they are over the following two years to adopt attitudes in which they value delaying initiation into romantic and sexual activity. In this case, we’re able to predict future attitudes, afteraccounting for current attitudes, by examining the degree of engagement within the teen-best friend interaction.