Measuring the usability of groupware applications with a model-driven method for the user interaction analysis

Abstract

The evaluation of the usability of an interactive system implies analyzing the degree in which the system can be used with effectivity, efficiency and satisfaction in a specified context of use. Nowadays these evaluations often approach the difficulty of measuring the usability of groupware applications that enable a group of users to interact between them in every place and at every time by means of mobile devices. The laboratory-based evaluations are far from being able to reproduce this real world with users geographically distributed and interacting all the time. However, the usability evaluations can take advantage of computational support that observes the behavior of final users in natural settings to calculate measurements with information about the effectiveness and efficiency of the user interaction (rate of errors, speed of performance, etc.). Thus, this paper describes an environment that enables the evaluators to specify models with the user actions to be analyzed and with a set of usability measurements to be calculated. Then, the environment automatically generates the computational support to perform the usability evaluation that has been modelled. Finally, the environment generates a report with the results of the evaluation and applies a statistical model to find the relationships between the measurement values and the subjective satisfaction of the user.