ABSTRACT
Strategies of information seeking (IS) about other people on social network sites have been examined in various interpersonal relationships. However, research has yet to account for whether, how, or why different strategies yield information of different quality. This study employs perceived warranting value and characteristics of a source as explanatory mechanisms to account for the relationship between strategies of IS and the perceived accuracy of information. An experiment randomly assigned people (N = 157) to engage in one of four IS strategies on Facebook for the purpose of reacquainting with another person. Trust in a source mediated the relationship between certain IS strategies and the accuracy of information obtained, but perceptions of warranting value did not.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Ashley M. Peterson (MA, The Ohio State University) is currently a doctoral student in the Department of Communication Arts & Sciences at The Pennsylvania State University.
Andrew C. High (PhD, The Pennsylvania State University) is an associate professor in the Department of Communication Arts & Sciences at The Pennsylvania State University.
Notes
1 Although there is no theory-driven reason to suspect that men and women differ in how they evaluate these characteristics of information, we acknowledge the disparity in gender in our sample. Accordingly, we reran our analyses with gender included as a covariate. There were no changes regarding whether or not findings were statistically significant.
2 There are benefits and drawbacks to applying alpha adjustments to the pairwise comparisons (e.g., Bonferroni, Tukey-Kramer, etc.; Hayes, Citation2009; Hewes, Citation2003; O’Keefe, Citation2003). For example, imposing a correction can result in a failure to observe important findings (O’Keefe, Citation2003). Ultimately, researchers should make correction-related decisions based on the conditions of their study. In particular, Bender and Lange (Citation2001) advise against corrections when analyses are exploratory. Because differences among strategies of IS are assumed to exist, yet the nature of these differences is unclear, we elected to report all pairwise comparisons using least significant difference tests. However, we understand the value of reflecting on our findings in light of an adjustment to the p-value. Given six pairwise comparisons, a Bonferroni correction to p < .05 = p < .008, and provides the significance of each pairwise comparison. All values in which p < .001 represent those comparisons that meet this threshold.
3 A complete list of indirect effects is available upon request.