Article Preview
Top1. Introduction
Usability tests are an important method to determine the usefulness of interactive systems and products in realistic settings and with real users: In usability tests, participants solve tasks that they would typically work on with a certain system. By observing the interaction, problems and difficulties can be determined and corrected in the software. Furthermore user acceptance and satisfaction can be measured.
Usability tests are often conducted in usability labs, which are equipped with specialized hardware and software for audio and video recording, mouse tracking, screen recording or eye-tracking analysis. Usability tests usually cover ‘normal’ use cases and conditions: It is observed what works well and what problems arise in a regular use situation. Supported by the analysis techniques named above, experimenters are able to gain manifold insights into user behavior and possible improvements of the software.
However, in the laboratory important context factors might not be as present as in the real situation or even suppressed altogether (cf. Greifenender, 2011), such as noise, presence of other people, interruptions, bad or bright lighting, special hardware etc. Imagine, for example, an electronic train ticket vending machine. People typically use such systems in a public situation, possibly in a hurry because the train is leaving shortly, pressured by others waiting in the queue. It is easy to imagine that testing a vending machine under such conditions will yield other results that in a quiet and relaxed usability lab. In many areas simulations are used to specifically test how users and systems perform under difficult conditions or in risky situations, e.g. when training pilots or staff of safety-critical facilities and equipment.
In this paper we investigate the use of stressors in regular usability tests (i.e., not especially regarding safety-critical systems) to find out how they possibly influence users’ performance and also their evaluation of the product or system they tested. To that end, we conducted an experiment to compare usability tests under regular laboratory conditions with a situation where several stressors were induced, such as time pressure, noise, and social pressure. The results of our study provide preliminary insights into the effects of stressors during usability testing.