How does Interest in a Work Task Impact Search Behavior and Engagement?

One goal of using simulated work tasks in interactive information retrieval (IIR) experiments is to create a more relevant and interesting search experience for study participants. However, there is not much guidance about how to identify interesting tasks or how interest in a task impacts search behaviors and experiences, which is the purpose of this study. In this study, we created eight work tasks and asked forty participants to rank these tasks from most interesting to least interesting before they came into the lab for an IIR experiment. During the experiment, we asked participants to conduct searches for the two tasks they ranked as most interesting and the two they ranked as least interesting. Participants completed pre- and post-search questionnaires to characterize their interests in the tasks and their search experiences, including engagement. Participants rated their interest, prior knowledge and search experience, and the relevancy of interesting tasks significantly higher than uninteresting tasks. They also predicted these tasks would be significantly less difficult to complete. Participants reported significantly greater engagement with interesting tasks and they spent longer completing these tasks. However, there were no significant differences in participants' search behaviors including number of queries issued, number of SERPs, or number of documents bookmarked. These results provide evidence that our method of assigning tasks to participants that would interest and engage them, at least cognitively, if not behaviorally, was somewhat successful. This method can be used by others conducting laboratory IIR studies.