Comparing QWERTY and Dvorak Keyboard Speed: a Pilot Study

Abstract

The comparison between QWERTY and Dvorak layouts has been a controversial for many years. It is well known that QWERTY layout was not designed for efficient English typing. Conversely, Dvorak layout was developed by considering the characteristics of English. Nevertheless neither layout has been demonstrated empirically to be superior. The main reason is that it is almost impossible to find people who are equally skilled in both layouts. Comparative studies measuring the typing speed on Dvorak against QWERTY have generally failed to remove biases arising from the unequal familiarity of users. In preceding research, we suggested an efficient methodology for screening ultimate typing speed for different key layouts. Because it uses the users familiarity with the generally-used layout, it naturally also has biases. However these biases can be easily quantified and corrected. We had shown this methods validity and effectiveness from the relatively large-scale experiments in mobile keypads with a wider range of users. We apply this screening method to QWERTY and Dvorak to figure out which layout is faster and how much which layout is more efficient. We built a web based system to gather a large amount of detailed typing data in small cost. This enabled the objective comparison between two layouts and also discovering characteristics of English typing. This information can be beneficial to develop new input devices or methods for various computing devices these days.