Psychology
Permanent URI for this communityhttp://hdl.handle.net/1903/2270
Browse
Search Results
Item Levels of Automation and User Participation in Usability Testing(Laboratory for Automation Psychology and Decision Processes, 2004-04) Norman, Kent L.; Panizzi, ManueleThis paper identifies a number of factors involved in current practices of usability testing and presents profiles for three prototype methods: think-aloud, subjective ratings, and history files. We then identify ideal levels to generate the profile for new methods. These methods involve either a human observer or a self-administration of the test by the user. We propose methods of automating the evaluation form by dynamically adding items and modifying the form and the tasks in the process of the usability test. For self-administration of testing, we propose similar ideas of dynamically automating the forms and the tasks. Furthermore, we propose methods of eliciting the user’s goals and focus of attention. Finally, we propose that user testing methods and interfaces should be subjected to usability testing.Item Usability Testing Of An Internet Form For The 2004 Overseas Enumeration Test: Iterative Testing Using Think-Aloud And Retrospective Report Methods(Laboratory for Automation Psychology and Decision Processes, University of Maryland, College Park, MD., 2004-06) Norman, Kent L.; Murphy, ElizabethAn Internet form for the U. S. Census Bureau’s 2004 Overseas Enumeration Test was evaluated in two rounds of usability testing. Participants were assigned to one of two conditions: Think-Aloud, in which they talked about what they were doing; or Retrospective-Report, in which they completed the form and then talked about their experience while viewing a recording. Participants also completed follow-up tasks. Sessions were video taped and logged. Round 1 testing identified 28 usability issues. Round 2 testing found that 13 of the issues had been resolved following design changes made to the interface. Round 2 testing identified 21 new and continuing usability issues. Results suggest that changes made to the interface increased the likelihood that respondents would be able to successfully complete the form. Task completion times in the think-aloud condition were only slightly longer than they were in the retrospective condition, while retrospective reports required a substantial amount of added time.