Examining the factors that influence selection choices within interactive search

Historically, human search behaviours have been studied through experimentation using static two-dimensional displays. In these experiments, searchers are tasked with finding target object(s) within visual scenes or amongst stationary distractor objects. However, the vast quantity of theoretical models and research findings fashioned from this approach have one crucial limitation; they overlook the role of physical interactions. The central purpose of this thesis was to address this missing link by conducting search experiments that were interactive in nature. Specifically, these experiments focused upon the factors that drive cognitive decisions regarding which objects or areas to interact with, and how exhaustively to search these before termination. The first set of experiments, presented within Chapter 2, investigated the role of the low prevalence effect within interactive search; a robust effect within visual search by which rare targets are often missed by the searcher. Results revealed standard low prevalence effects upon response accuracy. However, in contrast to visual search, this was not a result of reductions in search exhaustiveness. The second set of experiments, presented within Chapter 3, examined the role of effort on interaction choices. The results confirmed effort to be an extremely strong driver of attentional selection but also highlighted that searchers were often unwilling to terminate a search before having checked and revealed all possible visual information within the display. The final set of experiments, presented In Chapter 4, investigated the important confound of time within interactive search. In the prior experiments, it was unclear whether selection biases were a result of the increased time taken to interact with high effort objects. The findings confirmed that whilst effort was the primary cause, time should not be overlooked within interactive search. Furthermore, results again confirmed that participants were unwilling to leave visual information unchecked within trials. Overall, this thesis and the combined results highlight that the behaviours conducted during visual search are not a good approximation of those conducted during interactive search. Primarily, interactive search tasks are conducted more exhaustively than their visual search counterparts and interaction choices are heavily influenced by the effort and time required to interact with objects.