Quality in use : addressing and validating affective requirements

From a user's perspective, product quality has typically been equated to usability, where the primary focus is performance (i.e. efficiency and effectiveness) and user satisfaction. This view on quality has been adequate to validate the 'fitness for purpose' for many products. However, many other products, such as computer games or at-home-technologies, are not based solely on performance-based qualities. By defining and validating the quality of these non-performance-based technologies using the same measures as performance-based technologies fails to address a key quality - the ability of the product to create fun, enjoyment, or other user affective experiences. This thesis investigates system quality issues in non-performance-based technologies, primarily addressing affect. It explores what affect is in relation to system quality, how can affect impact quality in use, what product characteristics can aid in the creation of positive user affect, and how to validate whether a product successfully created the desired user affect. To begin, current knowledge about the relationship between systems development, quality in use, and affect is summarised. Concepts used synonymously with affect were studied (e.g. satisfaction, user experience, emotion). It was found that these are related to affect, but are not affect themselves. Following this, a broad survey asking users to identify product characteristics that create positive affect during use is reported on. Results showed that the system characteristics of learnability, feedback, curiosity, and user cooperation can each increase and enhance positive user affect experienced. If it is an aim for the system to create an affective experience in a user then this should be considered a requirement of the system; and as a requirement, it is necessary to validate that it achieves this aim. To this end, a laboratory study exploring psycho-physiology and cued-recall debrief as methods to evaluate user affect was conducted. Results revealed that both methods show promise for evaluating user affect. Cued-recall is capable of identifying specific user affects, but is limited because it relies on the user to report these affects. Physiological measures can objectively identify when a person is experiencing affect, but is limited because it is not possible to determine what affect is being experienced. Combined, these methods represent an optimal and viable evaluation method for user affect