Investigating differences in search behavior
MetadataShow full item record
Information search behavior (and decision making) is the focus of this dissertation. To this end, our goal is two-fold: a) explore the various antecedent variables to information search and measure the corresponding effect sizes; b) compare and contrast consumer information search behavior in three different modes of search (i.e., the medium in which information search is undertaken) – in-store, e-commerce and m-commerce. As a method for summarizing extant traditional search literature, we conduct a meta-analysis, with information search (i.e., “total amount of search” in traditional channels) as the dependent variable. 81 antecedent variables are uncovered from 65 studies, and the meta-analysis is carried out on 44 variables, 37 of which show significant effect size(s). Moderator analysis suggests that age, gender, product type and income are the most significant moderators of consumer information search. Next, two laboratory experiments are conducted with information search, evaluation of alternatives, purchase/decision and post purchase behavior (variously operationalized) as dependent variables. The first experiment is a 3x3 between subjects design (usable sample size 162), where each factor has three levels. Mode of search (the levels are in-store, e-commerce, m- commerce) is the first factor, and task type (simple Æ complex), is the second factor. The second experiment is a 2x3 mixed design (usable sample size 31), where mode of search is the within factor and task type is the between factor. The second experiment replicates the results of the first one. Competing predictions are made and different hypotheses are tested based on different theoretical frameworks (e.g., cost-benefit framework, categorization theory, media-richness theory). Our findings suggest that the amount of information searched in the different modes, follows the predictions made by cost theory (i.e., an inverted U-shaped curve). Lesser amount of search is undertaken in the most rich medium (i.e., in-store), while it increases as one moves on to e-commerce. However, it is the least in m-commerce (i.e., least rich medium). Further, “task-mode fit” is perceived, supporting the hypothesis that certain modes are more suitable for specific tasks than other modes.