Comparing task practicing and prototype fidelities when applying scenario acting to elicit requirements

Abstract

Identifying accurate user requirements early in the design cycle is of the utmost importance in system development. The purpose of this study of requirements elicitation was to compare the results of involving the user early in the design cycle using a low-fidelity prototype with the results of involving the user after a high-fidelity prototype was available. Three groups of potential users applied the method of Scenario Acting. Participants in the first group were given a working prototype of a human capital development system. The participants of the second group were given a detailed description of proposed features of the system and were told to practice on a paper prototype or with current methods, such as an Internet browser. These groups then practiced the tasks for some time before participating in the Scenario Acting. The third group received a brief description of the objectives of the system and did not practice the tasks. The results of the study showed that the use of the high-fidelity prototype was not helpful for eliciting requirements when working with users. However, the second group, taking time to practice the tasks given a low-fidelity paper prototype outperformed the others. Furthermore, the analysis of the Scenario Acting sessions revealed that two sessions were better than one, especially for participants of the group working with a low-fidelity prototype. An analysis of the topic of requirements showed that there was no difference between the groups on the domain tasks (here, human capital development), but the group practicing on the high-fidelity prototype commented more on its ease of use and usefulness than the other two. By comparison, the group practicing on low-fidelity prototype had more comments on the practice of the work and output of the tasks.