Abstract

Electronic reading opens new avenues especially with the advance of modern reading devices. The new generation of Personal Digital Assistants PDAs becomes more popular and more affordable. Therefore, while displays keep shrinking in size, it is needed to re-evaluate typefaces used in these devices as they form a substantial component in the reading field. In this research, a survey was conducted to identify Arab community preferences of 13 selected fonts on PDAs. Also, it inferred the popularity of using these devices for reading. From the participation of 53 subjects in this survey, it was deduced that e-reading using PDAs among Arab communities is increasing dramatically, which necessitates the need of investigation for better fonts used in these devices. Moreover, the results from font evaluation based on people preferences reduced the number of studied fonts to six for further examination.
Three experiments have been conducted to investigate six Arabic fonts on PDAs from the perspective of legibility and readability to come up with the best fonts. In all three experiments, 138 subjects participated doing i3arabi Test over iPad and iPad mini devices. Two experiments were done to evaluate the legibility of the selected fonts. However, due to the nature of Arabic language, it was difficult to apply the same methods used to test Latin fonts. A pilot study was done to understand the problem, and results supported the mentioned difficulty. Therefore, a novel method named M-Short-Exposure method has been proposed to investigate the legibility of isolated Arabic letters and connected letters. The results indicate Geeza Pro and
Uthman SH fonts yielded the best performance in the first and second experiments respectively. Then an integration result has been concluded for legibility experiments confirming Geeza Pro and Uthman SH as the most legible fonts to be used on PDAs.
In readability experiment, reading speed and comprehension questions have been used over running texts of the selected fonts to measure their readability. It has been found that there is no correlation between reading speed and comprehension factors. Though, the results provide Yakout Reg and Uthman SH fonts as the most appropriate fonts to be used on PDAs for e-reading.
Finally, Our findings provide the most legible and/or readable font(s) among the tested set. Moreover, some recommendations have been made on better use of legible and/or readable Arabic fonts for different purposes.