Modeling On-Line Search Behavior Using Alternative Output Structures

Abstract

The Internet provides a fantastic opportunity for users to obtain information for which they would otherwise have no access. However the existence of billions of web pages presents as much of a challenge as it does an opportunity. Surveys of search engine use have determined that they generally do not support efficient and effective searching. Part of this problem is that little consideration has been given to the nature of the search task in designing the output of most search interfaces. This study investigated alternate output structures for presenting the results of on-line search. Forty participants completed two search tasks using both the standard list output structure and an alternate tabular structure. The results showed that the tabular structure supported a wider variety of search strategies, including some that were not feasible with the standard list. The tabular structure also supported faster parsing and was preferred by users. The study provides evidence that consideration of multiple search strategies would enhance the design of search output structures.