Negative Ties in Action: Investigating the Emergence of Avoidance Ties Among Pupils

Adrian Toroslu, Eva Jaspers


Recently, interest in negative ties has increased with researchers looking into diverse topics such as inter-ethnic friendship and foe ties between pupils. This interest comes as no surprise since negative ties play an important role in classical theories such as balance theory and structural balance theory. However, much is still unknown about the dynamics of negative networks and the different antecedents which lead to negative tie emergence. Some studies link social status struggles to the creation of aggressive ties between pupils. However, little is known if such findings also hold for more moderate negative ties. This research builds upon earlier work and studies an avoidance network of pupils nominating others who they avoid having contact with (i.e. a directed tie). We argue that such a network is particularly suitable for studying negative tie dynamics because, for the average person, a negative sentiment is likely to result in in the creation of social distance (i.e. avoidance) between them and the person they hold negative sentiments for. Additionally, most people do not like open conflicts and, as a result, if they do not like someone, they end up avoiding them. Furthermore, especially for adolescents, who are the focus in this paper, aggression may also co-exist with friendship relations, but for avoidance this is less likely. We employ status considerations, balance theory, and homophily arguments to arrive at our hypotheses. First, status theory is applied to argue why pupils establish avoidance ties in the first place as well as to link status differences (measured as incoming positive ties) between pupils to avoidance tie emergence. Second, the theory of intergroup conflict as well as the mechanism of homophily are applied to argue that avoidance ties are more likely to emerge between natives and non-natives compared to within these two social groups. Third, structural balance theory is used to make predictions about avoidance tie emergence in specific triad formations as well as about the network as a whole. We argue that over time the overall network structure should become more balanced. This paper utilizes a three-wave dataset collected in the school year 2017-2018 at two Dutch high schools with a total of 228 first year pupils distributed over nine classes. This dataset is particularly useful to test to outlined mechanisms because it consists of first year high school students who typically did not know one another before entering high school. Therefore, it is possible to observe tie formation in its beginning stages. We consider directed, unsymmetrical ties among pupils within each classroom. The data is analyzed using longitudinal multivariate social network analysis with the RSiena package in the software environment R. This software allows for the estimation of stochastic actor-based models which are able to model the co-evolution of negative and positive tie networks while accounting for individual level variables.

← Schedule