This article investigates the influence of test duration on user fatigue and the reliability of user ratings in the context of subjective Quality-of-Experience (QoE) assessment. The goal is to provide empirically grounded guidance for the design of lab-based quality experiments, particularly as concerns the overall duration of test sessions. Since subjective user tests tend to be time-consuming and costly, aspects of user workload and fatigue are relevant as they relate to a fundamental challenge: how to maximize test duration without compromising results quality by overly exhausting test participants? In order to address this challenge, we investigate the relationships between test duration, user fatigue, and rating behavior. Our analysis is grounded on measurements and observations made during three typical QoE lab studies with mixed audio, video, and web task profiles that assessed the impact of different network conditions on perceived quality. We measured participant workload and fatigue in two complementary ways: subjectively by means of a questionnaire and objectively by performing physiological measurements in terms of eye blink rates (EBR) as well as Electrocardiographs (ECG). Our results show that even after 90 minutes of active testing, participants’ quality gradings were still reliable despite the presence of measurable signs of fatigue. Thus, for comparable QoE lab user experiments, we recommend to stay within this limit in order to achieve a good balance between results quantity and results quality.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.