| Online social bookmarking systems allow users to assign tags (i.e., keywords) to represent the content of resources. Research on the effects of target audience on social tagging suggests that taggers select different tags for themselves, their community (e.g., family, friends, colleagues), and the general public (Panke & Gaiser, 2009; Pu & Chang, 2009). This experiment investigated whether the common ground theory proposed by (Clark, Schreuder, & Buttrick, 1983) could explain tag selection and assignment: whether knowledge presumably shared by the tagger and the target audience influences tag selection and assignment. Ninety subjects studying finance and journalism were randomly assigned to one of three groups. Each group then tagged the same set of 21 resources for a different target audience (i.e., self, colleague, or stranger). To explore the influence of shared knowledge on subjects' tagging behavior, the experiment addressed four questions: Does the number of tags assigned to resources vary across target audiences? Does the number of assigned tags appearing in tagged resources vary across target audiences? Does the number of concepts represented by assigned tags vary across target audiences? and Does the specificity of tags assigned to resources vary across target audiences? Analysis of tagging data showed that only the number of tags assigned to resources differed significantly between groups tagging for different target audiences. However, tag specificity and the number of tags appearing in the text of tagged resources showed patterns of tagging behavior suggesting the influence of common ground theory on tag production despite the lack of statistically significant differences. While findings for the number of tags assigned support the influence of knowledge presumably shared by the tagger and the target audience, research using larger samples is necessary to investigate the explanatory power of common ground theory for other aspects of tagging behavior. |
