August 31, 2011

Gossip makes the heart grow fonder: Relational aggression and friendship quality.

Banny, A.M., Heilbron, N., Ames, A., & Prinstein, M.J. (2011). Relational benefits of relational aggression: Adaptive and maladaptive associations with adolescent friendship quality. Developmental Psychology, 47 (4), 1153-1166.

This summary is a little different from previous posts, mainly because it is longer than I usually post. There are two reasons for this: First, there are two studies reported in the article, and second I really enjoyed reading the paper and want to get across why I think it is so interesting! If anything in the summary is confusing or unclear, please be sure to post a comment and I’ll soon respond.

As we have covered in previous posts, relational aggression involves behaviour which isn’t physically based; rather, it is aggression which aims to harm another person’s relationships and/or social standing. These authors note that this has been investigated in terms of its negative outcomes, but that less is known about its benefits. This is an interesting angle to take – after all, it is sensible to assume that using aggression has benefits for children  otherwise there would be little point in using it. If all aggression were maladaptive (only had negative outcomes for the aggressor) we would expect it to dissipate over time.

These authors were investigating relational aggression, friendship quality (both positive and negative). They note that relational aggression has been associated with both positive (intimacy, support) and negative (conflict, criticism, dominance) aspects of friendship in other research (though there were limitations to those studies, which I won't go into here). They suggest that some children may be attracted to dominant and relationally aggressive children because of a ‘halo’ effect i.e. some of the dominant child’s social standing can rub off on them. Intimacy is an issue which may be especially closely associated with relational aggression, reflecting the possible benefits of gossiping. What are these benefits?
  • Well, gossiping about others may increase intimacy in a friendship foster intimacy because sharing (someone else’s) private information suggests that the friend can be trusted with it (“wow, they must really trust me to tell me that”). 
  • If a relationally aggressive child is thought to be ‘putting themselves out there’ i.e. going out on a limb by spreading malicious gossip and opinions, then this may increase feelings of solidarity (“they’re saying what I’m thinking... hey, we really are alike”).
  • Excluding others and spreading gossip may also help young people to feel like they are part of a group and to promote a feeling that there is a close bond among friends.

Relational aggression can, of course, also lead to problems. For example, exclusion of others may become exclusivity in a friendship and may be linked to feelings of envy or jealousy.

These authors designed two very different studies to examine these issues.

Study 1:
This study was designed to look at the longitudinal relations between relational aggression and friendship quality (positive and negative) in stable, reciprocal friendships. This means that the authors were looking to see how relational aggression at one point in time predicted friendship at a second point of time. They were able to take into account friendship quality at the first point in time too. “Stable, reciprocal friendships” were friendships where two young people nominated each other as a best-friend on both occasions when data were collected.

Analyses were based on a total of 62 adolescents (58% female) who were in Grades 6, 7, and 8 at the start of the study (i.e. aged approx 11, 12, and 13). These were a subset of a larger data set of 520 students who all took part at two points in time (time 1 and time 2 were separated by 11 months). The subset were the stable, reciprocal friendships. At both points in time, participants had to nominate who in their class were their closest friends and who was their very best friends. They also nominated who in their class was relationally aggressive, and also who was overtly aggressive (this latter measure was taken because the two types of aggressive tend to go hand in hand so it is important to statistically ‘remove’ the effects of overt aggression when conducting the analysis). Friendship quality was assessed using self-report scales.

Comparing boys and girls, girls had more positive friendship qualities. No other gender differences (on overt or relational aggression, or on negative friendship qualities) were significant.

The finding from the longitudinal data was both simple and potentially very telling: the more relationally aggressive a child was at Time 1, the more they considered the positive qualities of their friendship to improve at Time 2. But here’s the rub: the relationally aggressive child’s friend did not think the positive friendship qualities had improved. So, relationally aggressive children saw their best-friendships are getting better over time, while their best-friends did not see the same increase in positive qualities (note that the best friend did not report a decrease either).

There were no relationships for negative friendship qualities – being relationally aggressive made no difference to later negative outcomes. It is important here to remember that these results are for adolescents in enduring reciprocated best friendships. A potential criticism is that this group may be characterised by lower levels of relational aggression anyway, and so we might expect that their negative friendship qualities would not be affected (or be affected less so) than young people in less enduring relationships. The authors note that the reciprocal friendship group did display lower levels of relational aggression than those not in longitudinal reciprocated best friendships – however, levels of relational aggression did not predicted whether a friendship would dissolve or not.

Study 2
This second study really took me by surprise – actually observing aggressive interactions is rarely done in research on children and young people’s aggressive behaviour because of the practical difficulties involved. However, here they used observational methods to examine how “relationally aggressive talk” (e.g. saying negative things about others in a gossipy way, trying to change the relationship between who is there and who is not) might link to friendship quality. The authors focussed on this form of talk because they are the types of behaviours that other young people might not be able to see (and so might have been less likely to be reported in Study 1).

In this second study, there were 56 adolescents (47% female) in Grades 9 and 10 (about 14 to 16 years old) who asked a best friend to also take part. They attended a laboratory session where they completed a questionnaire (assessing friendship quality) and took part in an activity which was recorded. The target participants (those who invited someone else to take part) also took part in a telephone interview 6 months later (again assessing friendship quality). In observations, the pairs asked to discuss a number of things for 5 minutes each (e.g. planning a party). Of particular interested was the section where they had to talk about their peer group. Relationally aggressive behaviours were assessed by looking at the recordings.

As in the Study 1, girls had higher positive friendship qualities and there were no other gender differences.

The authors then took a different approach to the one taken in Study 1. They compared  friendships where either
  1. both members said the other was their ‘best friend’ at both time points, or 
  2. one said ‘best friend’ and the other said ‘close friend’ at one of the time points. 
They were able to do this because they asked the young people taking part to say, in private, whether their partner was a ‘best’ or ‘close’ friend. The extent to which the target child engaged in relationally aggressive talk at time 1 predicted increases in positive friendship quality at time 2 – however, this was only true for the group (1). For group (2) relationally aggressive talk at time 1 was not related to negative friendship quality at time 2. I thought this was really interesting – it suggests that the positive benefits of engaging in relational aggression may be restricted to very close friendships. It was also interesting to see that in both of the studies reported here relational aggression did not harm friendships quality it seemed to be a no-loss situation where the young person would either improve the quality of their friendship, or there would be no impact.
Of course, there are some limitations to the study, most obviously a relatively small sample size (though a decent sample was recruited to begin with - it is just that choosing those with reciprocal friendships means you have to cut down the sample available to analyse). But the strengths, including the observational methods, peer-nomination methods, and longitudinal design, all give us some confidence in the findings.


  1. i like the focus of this article b/c it acknowledges the benefits of relational aggression.

    presumably, such behaviour is not problematic for those who are perpetrating it (if it brings them closer with their friends, etc.). but what about those who are the targets? did the article include any discussion of relational aggression from this perspective?

    i also thought it was intersting that in the first study, one of the children was identified as "relationally aggressive", while their partner was not. isn't it more likely that both of them would've been engaging in this behaviour?

  2. Thanks for your comments.

    Certainly, the targets of aggression get the raw end of this deal! But there are quite a lot of articles already looking at the effects on victims (almost universally negative, as you might expect), so this article was considering instead what the aggression achieves for the perpetrator (assuming that the perpetrator isn't simply a psychopath... which is a pretty good assumption to make since the behaviours referred to are so widespread).

    I'm not sure about your second point - I don't think the first study mentioned anything about one friend being aggressive and the other not. Perhaps I've summarised something poorly though? Can you copy-n-paste what I said that made you think that? I think you're probably right - and in fact the authors analysed their data using a new method (called APIM analysis if you're interested!) that takes into account the fact that the pairs of young people are likely to behave in similar ways due to the fact that they are a 'unit' in some sense (i.e. friends here).

    Thanks again for your comments.

  3. from study 1: "the relationally aggressive child’s friend did not think the positive friendship qualities had improved" - which implied that one child was relationally aggressive, while the other was not.

    i appreciate for your answer.

  4. Sorry - I wasn't clear enough. The article was looking at specific children (called 'actors') and the authors used each actor's level of relational aggression at time 1 to predict the actor's perception of friendship quality at time 2. The authors also collected information on each actor's best friend (called 'partners'). The partners also reported on friendship quality at times 1 and 2. Nothing is actually reported in the article about the partners' levels of relational aggression.

    so - my comment just means that the 'partner' didn't think that their friendship with the relationally aggressive child had been improved between time 1 and time 2. It is not meant to imply that the partner was not a relationally aggressive child.

    ...any clearer?!