HARMONY AND TENSION ON SOCIAL NETWORK SITES

Abstract

The need to maintain harmony among one's social contacts is proposed in this paper to impose constraints on the interconnectivity between users of social network sites (SNS). A particular focus is on the connectivity between different social spheres. It is hypothesized that the type and number of social spheres and technological features of SNS interact such that increased levels of social tension result. These ideas are supported by the findings from a survey study among Facebook users. Social diversity of the Facebook network predicted online tension as did the number of family members on Facebook, in contrast to work and social contacts. Furthermore, evidence was found to support the idea that tension might impose an upper limit on network size. Follow-up interview data also showed that online tension was predominantly about unwanted connectivity between the spheres. All the technological features that users reported as problematic focused on the easy access to and broadcast of text and pictures. Findings are discussed in light of unintended, negative side-effects of SNS and social media in general.

Keywords:

• social network sites
• social sphere
• balance theory
• network diversity
• network maintenance
• social circle

Acknowledgements

This research was jointly funded by the EPSRC, UK and ESRC, UK.

Notes

To control for the skewed distribution, we repeated parts of the analyses using a logarithmic transformation of tension. Effects remained essentially unchanged.

The regression analyses carried out for work, social and family contacts separately showed highly similar results. We further tested whether online tension merely follows from offline tension and is not specifically related to the use of SNS technology. The analyses were, therefore, repeated using the corresponding offline network figures (number of family, work and social contacts as well as diversity) as predictors. No significant findings for these offline spheres emerged.

Note that the Blau index is sensitive to the proportionate representation of social categories. We also used a raw count of categories that might capture other relevant aspects of diversity here. The results, however, were virtually the same.