Thesis title: WEBSITE COMPLEXITY AND USABILITY: THE ROLE OF MENTAL WORKLOAD
The evolution of digital interfaces has changed the way people perform their main daily activities (Parameswaran & Whinston, 2007; Wattal et al., 2010). The spread of the first personal computers, between the second half of the 1970s and the first half of the 1980s, marks the beginning of the digitisation process. The first machines were complex computing systems that only experts were able to use. With the “revolution” of graphic interfaces and the introduction of the paradigm “What you see is what you get” (Goldberg, 1988; Myers, 1984; Smith et al., 1982), there was a rapid spread of computers and digital interfaces for non-expert users. All individuals today interact with web-based interfaces and systems for most of the day, both at work and in their private life. This technological development, which involved the transition from physical to graphical interfaces, has been accompanied and guided by the development of a multidisciplinary scientific discipline known as “Human-Computer Interaction” (Grudin, 1990; Preece et al., 2015).
The term “Human-Computer Interaction” (HCI) generally refers to the study of the design, evaluation and development of technological interfaces (Hewett et al., 1996). HCI benefits from the contribution of disciplines such as engineering, computer science, cognitive psychology, sociology and anthropology (Carroll, 1997; Rogers, 2004; Lazar, Feng & Hochheiser, 2017). The main research areas of HCI include the study of cognitive processes involved during human-machine interaction, the development of principles, guidelines and heuristics to be applied in the design and evaluation of interfaces (Preece et al., 2015). Cognitive psychology has strongly contributed to the development of human-centred interfaces. HCI has integrated the main assumptions of cognitive models in the design phases of digital instruments and devices (Rogers, 2004). Some noteworthy examples refer to the studies on memory conducted by Baddeley (1992) and Miller (1956) or the tripartite theory of Atkinson and Shiffrin (1968). These studies have primarily influenced how human-machine interaction models are conceived and implemented. (Card et al., 1983). The main objective of the researchers has been to reduce the cognitive load resulting from individual- interface interaction (Mandel, 1997; Preece et al., 2015). Several studies have shown that when cognitive demands are either too high, or too low, these negatively affect the performance and user experience (Xie and Salvendy, 2000).
Mental workload is crucial in the design and evaluation of digital interfaces. Digital system designers must be able to predict the mental workload imposed by the system. Paying attention to users' mental workload is essential to enable them to use digital systems with effectiveness, efficiency and satisfaction, taking care to improve “perceived usability” and decrease user “mental workload”.
Research in this area has produced numerous studies and identified different metrics capable of estimating the mental workload experienced by a user during the execution of a task (O’Donnel & Eggemeier, 1986; Wierwille & Eggemeier, 1993). However, we are still far from being effectively integrated into usability assessments. This deficiency in the reference literature is mainly due to the fact that the constructs of mental workload and usability have been investigated in different fields of application and with reference to different types of users.
HCI Researchers (see Carrillo et al., 2017 for review) distinguished “inexperienced” or “occasional users” by “operators” and “expert users”. While the first ones have been the focus of User eXperience (UX) studies, the second ones have been more involved in Human Factor (HF) research. Research involving occasional users focused on the user experience that emerges from the interaction with an interface. The topics of greatest interest in UX research are “User satisfaction”, “Ease of use”, “Consistency”, “Affordance”. Researchers focus their efforts on designing intuitive and easy-to-use interfaces that support the user in achieving their goals. In this field, the concept of “usability” is fundamental and generally refers to aspects related to the quality of a system. Human Factors studies on operators and expert users, on the other hand, focused on the human-machine interaction in “high complexity” work environments (e.g., aviation, aerospace, healthcare, etc.) where specific skills and knowledge are required to use certain interfaces.
HF deals with psychological and physiological aspects of human capability, able to influence the interaction between the operator, the systems and the procedures of his working environment, and to directly affect the outcome of events. The high safety standard required in domains, such as aviation and healthcare industry, led to a wide exploration and application of HF theories and practices to the design of services, system interfaces and procedures. Some of the topics of greatest interest in HF research are “human error”, “situation awareness”, “automation” and the “mental workload” experienced by an operator who interacts with a system.
This compartmentalised division has the consequence that for a long time, there has never been any mention of “mental workload” in usability studies.
However, the concept of mental workload could provide important guidance for both the design phase of a system and the evaluation and improvement of its usability. Identifying reliable metrics that can provide objective information on these phenomena as well as their relation remains an open challenge for the scientific community. However, in this regard, eye movements analysis is a promising measure. Eye-tracking has gained increasing popularity in the scientific community as a technique to evaluate usability and mental workload (May et al., 1990; Jacob & Karn, 2003; Pan et al., 2004; Poole & Ball, 2006; Majaranta & Bulling, 2014). The analysis of eye behaviour (e.g. fixations, saccades and scanpath analysis) allows researchers to obtain real-time information on user- interface interaction. Moreover, this technique has high ecological validity and returns objective measurements.
Considering this brief premise, the purpose of this work is to contribute to the investigation of the relationship between mental workload and usability in relation to the use of digital interfaces. The constructs of usability and mental workload, and the main subjective and objective metrics used for their evaluation will be illustrated. A particular focus will be made on the eye movements recording techniques used for the usability and mental workload evaluation. In the last part of this work, some experimental studies will be presented. The structure of the manuscript consists of four parts. In the first part, the concept of usability will be described. The most widely used definition in the literature is reported, and the basic principles of usability are explained. Considerable attention is devoted to the different techniques used for its measurement. The differences between formative assessment methods and summative or comparative assessment methods (Scholtz, 2004), techniques involving users and heuristics-based techniques will be explained. In addition, the main subjective usability questionnaires will be described in detail.
The second part will focus on the mental workload construct. After illustrating the main definitions and theories behind this construct, the various techniques useful for its measurement will be explained. In particular, the manuscript will focus on the most used physiological and subjective measures in the reference literature. The third part provides an examination of the use of eye movements in psychology, the main characteristics of eye behaviour, the metrics and recording methodologies used. Also, this part of the work focuses on the most widely used ocular metrics for a measure of usability and mental workload, highlighting their differences and convergences.
In the fourth part, different experimental studies will be presented. This section will deal with some research questions accompanied by the results of the relative experimental analyses. The first study had the objective to validate a new tool for a quick and reliable estimation of the usability perceived during the use of digital interfaces. The comparisons between the new tool and other usability questionnaires will be analysed. The second and third studies investigate the relationship between usability perceptions and mental workload associated with browsing complex websites. A promising metric based on the analysis of eye movements, the NNI, was used in these studies. This metric provides real-time information about the mental workload experienced by a user when using an interface. Although the results obtained are encouraging, defining these comprehensive studies would be ambitious. In conclusion, the limitations that emerged during the studies are discussed, and ideas for future research are proposed.