A Study of Usability of Arabic Forms for Smartphones

Abstract

Researchers have established a concrete and solid ground for web forms de- sign and their usability testing since the nineties. However, since the dramatic shift towards devices that carries a varied set of properties, such as, smartphones and tablets, developers started to build forms to suit these devices. As a conse- quence, a radical shift in research and usability testing needed to strike as well. We attempt to work with Arabic forms that are designed particularly for mobile devices (smartphones to be exact). Our goal, was to come up with empirically tested usability guidelines for Arabic mobile forms through achieving a set of objectives. Firstly, we have investigated the e ect of smartphones screen size on error rate. In other words, do smaller screens increase the number of errors made by the users during form lling tasks? Secondly, we have addressed the following question: what is the best location and timing for an error message to appear on smartphones forms? Finally, we have examined the use of graph- ics (highlighting the erroneous eld) on form lling. The above issues as come to our knowledge never been addressed before for Arabic smartphones' forms. We have reached that conclusion after a detailed literature review. The above matters were measured in this dissertation in terms of e ciency, e ectiveness, and satisfaction. Our experiments involved two main tasks: questionnaire dis- tribution and form lling. Our ndings indicated the existence of a negative relationship between screen size and number of errors made by the user in form lling, users are less likely to make errors when the screen is big. Our ndings have shown that the best timing for an error message to appear is before sub- mission (immediate feedback). This timing category involved less time needed to complete the forms and correct errors, less errors made by the users, and it was voted by the participants as the best timing. As for the best location for the small screen, our ndings revealed that below text box is the best location for a message to appear, while for the big screen, below label is the best location for a message to be shown. Above label messages scored the worst which in turn implies that placing the error messages above the label is considered a bad news when it comes to error presentation. Our last experiment has shown that the use of graphics (highlighting the erroneous eld) is very helpful. It captures the user attention instantly which implies the users ability to identify the location of the errors faster and much better.