Researchers Tulin et. al conducted a study with aims to investigate whether structural properties (group size and density) can predict perceived cohesion in small social groups. Tulin et. al’s review on the literature concerning group cohesion showed a common trend, in that it was often the case that emphasis is placed on individual perceptions, which may not reflect the actual social structure of the social groups. This study took a different turn and drew upon on social network theory to examine the relationship between structural group characteristics while simultaneously taking into account individual perceptions of group cohesion. The researchers did this by leveraging Facebook data, extracting and partitioning the social networks of 109 individuals into varying groups. Preceding this they surveyed the participants, asking questions related to their perceptions of cohesion, and computed group density and size using social network analysis.

The fundamental research question of this paper was whether or not the structural properties of groups can predict individual perceptions of cohesion. Tulin et. al presented two hypothesis:

H1: Group density is positively related to perceived group cohesion

H2: Group size is negatively related to perceived group cohesion.

Group density is defined as the number of observed connections divided by the number of possible connections in a group. The researchers make argument that group density and group size are linked both theoretically and mathematically such that larger groups tend to be lower in density. As group size increases linearly, the potential number of possible ties also increases. To maintain the same level of density, larger groups on average need to establish many more connections between their group members than lesser populated groups. Tulin et. al argue that this would be limited based on time and cognitive constraints, leading to the belief that smaller groups are both more manageable cohesive.

The sample population of this study consisted of One hundred and nine participants (37 men, 72 women) who were recruited on the campus of a European University. Facebook membership was a requirement of participation due to the nature of the study. The nodes in this network were the participants and their friends on facebook. The edges were the actual relationships between the Facebook users and their friends. Data was collected from participants using 3 surveys:

Perceived cohesion survey:

Tulin et. all used an adapted version of a perceived cohesion scale that originally developed for sociological research on cohesion in large reference groups, such as colleges or cities.  The scale consisted of six items and measures which perceived cohesion on two dimensions: the individual’s sense of belonging and feelings of morale. This was scored on a 7-point Likert scale.

Group cohesiveness survey:

A group cohesiveness scale was also used which consisted of seven items scored on a 5-point Likert scale. Questions that were asked focused on feelings of acceptance and participation. At the bottom of the group cohesiveness scale an open-ended question was included in order to identify the type of group that was being investigated i.e.: family, friends from home, university students, sports team etc.

Facebook survey:

The final set of questions the researches used assessed participants attitudes towards Facebook in general and privacy precautions that they use. Items were scored on a 6-point scale.

Study results did not support the researchers first hypothesis. According to the data, density was not linked to perceived cohesion. However size had a small negative effect on perceived cohesion, which Tulin et. al believe suggests that people may perceive smaller groups as being more cohesive.



Tulin, M., Pollet, T. V., & Lehmann-Willenbrock, N. (2018). Perceived group cohesion versus actual social structure: A study using social network analysis of egocentric Facebook networks. Social Science Research, 74, 161–175.