Is Less Sometimes More? An Experimental Comparison of Four Measures of Perceived Usability.

Abstract

In usability studies, the subjective component of usability, perceived usability, is often of interest besides the objective usability components, efficiency and effectiveness. Perceived usability is typically investigated using questionnaires. Our goal was to assess experimentally which of four perceived-usability questionnaires differing in length best reflects the difference in perceived usability between systems. Conventional measurement wisdom strongly favors multi-item questionnaires, as measures based on more items supposedly yield better results. However, this assumption is controversial. Single-item questionnaires also have distinct advantages and it has been shown repeatedly that single-item measures can be viable alternatives to multi-item measures. N = 1089 (Experiment 1) and N = 1095 (Experiment 2) participants rated the perceived usability of a good or a poor web-based mobile phone contract system using the 35-item ISONORM 9241/10 (Experiment 1 only), the 10-item System Usability Scale (SUS), the 4-item Usability Metric for User Experience (UMUX), and the single-item Adjective Rating Scale. The Adjective Rating Scale represented the perceived-usability difference between both systems at least as good as, or significantly better than, the multi-item questionnaires (significantly better than the UMUX and the ISONORM 9241/10 in Experiment 1, significantly better than the SUS in Experiment 2). The single-item Adjective Rating Scale is a viable alternative to multi-item perceived-usability questionnaires. Extremely short instruments can be recommended to measure perceived usability, at least for simple user interfaces that can be considered concrete-singular in the sense that raters understand which entity is being rated and what is being rated is reasonably homogenous.