Suicide and Friendships: The Clustering Coefficient
On Page 45 of Networks, Crowds, and Markets, Professor Easley refers to a study by Bearman and Moody where the two researchers found that teenage girls with a low clustering coefficient has a higher rate of suicidal thoughts than others. The suicide rate of adolescents in the US has been rising dramatically in the past decade. Research has already identified several basic risk factors: depression; exposure to suicide or suicide attempts by family or friends; substance or alcohol abuse; and having guns in the home. Bearman and Moody suggest that in addition to these factors, network intransitivity is another significant predictor for adolescents’ suicidality, especially for female teenagers.
In this study, Bearman and Moody analyzed friendship data on 13,465 adolescents and studied the relationship between social connections and suicide ideation. They identified two relations among teenagers: transitive and intransitive. If A and B are friends, and B and C are friends, then A and C are “two friendship steps apart.” If these two are also friends, then the resulting triad (A, B, C) is “transitive.” This concept of transitivity was discussed in our class as well when we learned about the Triadic Closure Property. If A and C are not friends already, then different factors such as opportunity, trusting, and incentive are likely to draw those two together and eventually make them friends. If this happens, B’s “clustering coefficient” rises.
When the clustering coefficient rises, transitive relations are formed. In these relations, friendship groups are closed and dense, and individuals within those groups are all friends to each other. Intransitivity would mean the opposite, where a person may be friends with multiple disconnected people. When this was measured among the study subjects, girls with a low clustering coefficient were significantly more likely to contemplate suicide than those with a higher clustering coefficient.
Although the results of the study are troubling, a few conclusions can be made here. First, it seems that close friendship groups are more important to girls than to boys. Although both girls and boys were affected by their friendship environment, girls were more likely to contemplate suicide than boys when they lacked a transitive group of friends. More importantly, school communities need to be active in analyzing friendship data among their students and do their best in making sure that nobody is left alone. This may be a key to preventing adolescent suicide, the third leading cause of death among adolescents today.