Web applications have become the backbone of business, information exchange, and social networks. In this kind of applications, usability is considered to be one of the most important quality factors, since the ease or difficulty that users experience with this kind of systems will determine their success or failure. However, there are several shortcomings with the existing Web usability evaluation approaches such as: the concept of usability is only partially supported; usability evaluations are mainly performed when the Web application has been developed; the lack of guidelines on how to properly integrate usability into Web development, and the shortage of Web usability evaluation methods that have been empirically validated. In addition, the majority of Web development processes do not take advantage of the software artifacts produced at the design stages. These intermediate artifacts are principally used to guide developers and to document the Web application but not to perform usability evaluations. Since the traceability between these artifacts and the final Web application is not well-understood, performing usability evaluations on these artifacts can be difficult. This problem is alleviated in Model-Driven Web Development (MDWD) processes in which intermediate artifacts (models), which represent different views of a Web application, are used in all the steps of the development process, and the final source code is automatically generated from these models. By considering the traceability among these models, their evaluation allows usability problems which would be experienced by end-users of the final Web application to be detected, and provides recommendations to correct these usability problems during the earlier stages of the Web development process. 

This PhD thesis aims to contribute towards addressing the aforementioned limitations by proposing a usability inspection method that can be integrated into different Model-Driven Web development processes. The method is composed of a Web Usability Model that breaks down the concept of usability into sub-characteristics, attributes and generic measures, and a Web Usability Evaluation Process (WUEP) that provides guidelines on how the usability model can be used to perform specific evaluations. The generic measures from the usability model must be operationalized in order for them to be applied to the software artifacts of different Web development methods and at different abstraction levels, thus allowing usability to be evaluated at several stages of the Web development process, especially during the early stages of development. Both the usability model and the evaluation process are aligned with the latest ISO/IEC 25000 standard for software product quality evaluation (SQuaRE). 

The proposed usability inspection method (WUEP) has been instantiated in two industrial model-driven Web development methods (i.e., OO-H and WebML) in order to show the feasibility of the approach, and WUEP has also been empirically validated by means of a family of experiments in OO-H and a controlled experiment in WebML. The objective of our empirical studies was to evaluate the participants’ effectiveness, efficiency, perceived ease of use and perceived satisfaction when using WUEP in comparison to a widely-used industrial inspection method: Heuristic Evaluation (HE). The statistical analysis and meta-analysis of the data obtained separately from each experiment indicated that WUEP is more effective and efficient than HE in the detection of usability problems. The evaluators were also more satisfied when applying WUEP, and found it easier to use than HE. Although further experiments must be carried out to strengthen these results, WUEP has proved to be a promising usability inspection method for Web applications which have been developed using model-driven Web development processes.