Ties, Likes, and Tweets: Using Strong and Weak Ties to Explain Differences in Protest Participation Across Facebook and Twitter Use

Abstract

Based on the theoretical concepts of social networks and technology affordances, this article argues that different social media platforms influence political participation through unique, yet complementary, routes. More specifically, it proposes that Facebook and Twitter are conducive to protest behavior through two distinct mechanisms: whereas the influence of Facebook use is more effective through communication with strong-tie networks, the impact of Twitter use is more effective through communication with weak-tie networks. To test these expectations, we analyze data from a cross-sectional, face-to-face survey on a representative sample of Chilean youths conducted in 2014. Findings in the study lend empirical support for these hypotheses. Consequently, while different social media (in this case, Facebook and Twitter) are similar in their participatory effects, the paths through which this influence occurs are distinct, a finding that highlights the importance of studying political behavior across different media platforms.

Keywords:

• political behavior
• protest
• social media
• social networks
• strong ties
• weak ties

View correction statement:

Corrigendum

Funding

The first author received funding from Chile’s National Commission for Scientific and Technological Research (CONICYT) through grants CIGIDEN/Fondap/15110017 and Fondecyt Iniciación/11140897. The second author received funding from CONICYT grants Fondecyt Regular/1140061 and Fondecyt Regular/1170324.

Supplemental Material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/10584609.2017.1334726.

Notes

1. Following international trends (e.g., Duggan, Ellison, Lampe, Lenhart, & Madden, 2015), the frequency of use is skewed for both Facebook and Twitter, which may be problematic when employing statistical techniques that assume normality (see histograms in the supplemental Appendix). Considering the cutoff point of ±2.0 as the acceptable range of skewness for a “normal” distribution, the three Facebook measures are “normal,” but the three Twitter measures are “skewed.” Consequently, we replicated the analysis using three different approaches: (a) transformation (i.e., using the square root function, to make the distribution of the Twitter measures more symmetrical); (b) dichotomization via median split; and (c) categorization (i.e., creating indicator variables for each level of the Twitter measures). In all three cases, the results are the same in substantive terms to those reported in the main text. Thus, the analyses reported here employ the original measures.

Additional information

Funding

The first author received funding from Chile’s National Commission for Scientific and Technological Research (CONICYT) through grants CIGIDEN/Fondap/15110017 and Fondecyt Iniciación/11140897. The second author received funding from CONICYT grants Fondecyt Regular/1140061 and Fondecyt Regular/1170324.

Notes on contributors

Sebastián Valenzuela

Sebastián Valenzuela is Associate Professor, School of Communications, Pontificia Universidad Católica de Chile.

Teresa Correa

Teresa Correa is Associate Professor, College of Communication and Literature, Diego Portales University.

Homero Gil de Zúñiga

Homero Gil de Zúñiga is Medienwandel Professor, Department of Communication, University of Vienna & Research Fellow, College of Communication and Literature, Diego Portales University.