Off-campus UMass Amherst users: To download campus access theses, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Theses that have an embargo placed on them will not be available to anyone until the embargo expires.

Access Type

Open Access

Document Type


Degree Program

Industrial Engineering & Operations Research

Degree Type

Master of Science in Industrial Engineering and Operations Research (M.S.I.E.O.R.)

Year Degree Awarded


Month Degree Awarded



Soft keyboard, position information, audio feedback


Soft keyboards have become ubiquitous, especially with the introduction of the iPad. This study aims to determine for experienced touch typists whether there are characteristics of soft QWERTY keyboards that can make them easier to use and why those characteristics provide an advantage. Two characteristics would appear to be of central importance. First, hard keyboards provide home row positioning information that is not as easily provided by soft keyboards. Second, hard keyboards also provide auditory and tactile feedback when a key is depressed, something not generally provided with soft keyboards.

In order to test the hypothesis that the absence of home row positioning and key strike feedback information can reduce expert touch typists’ speeds on soft keyboards, expert touch typists were run in two experiments. In Experiment 1, soft and hard keyboards in landscape and portrait mode were evaluated. The hard keyboards had the standard home row positioning and key strike feedback whereas the soft keyboards had neither. If these are important elements in typing speed, then experienced hard keyboard typists should type less quickly when using soft keyboards than when using hard keyboards. Moreover, if reducing the footprint of the keyboard, from landscape to portrait, requires more eye movements, then typists using both hard and soft keyboards should be slower when using the portrait size keyboard than when using the landscape size keyboard. Perhaps not surprisingly, experienced hard keyboard touch typists do less well when entering information on soft keyboards without home row positioning information or auditory feedback. Moreover, both groups appear to type more slowly in keyboards laid out in a portrait format than they do in keyboards laid out in a landscape format.

In summary, the results from Experiment 1 suggest that both home row positioning information and auditory key strike feedback should speed performance. In Experiment 2, an attempt was made to determine just how much of a gain can be made in the typing speed of more experienced soft keyboard users if home row positioning information (tactile feedback), auditory feedback, or both are added. Participants were run in four conditions: auditory key strike feedback (with and without) was crossed with tactile home row positioning information (with and without). Participants included expert level hard keypad QWERTY touch typists who have had at least five hours’ typing experience with an iPad. Participants were given four passages to type, all of equal length and all balanced for letter frequency. Participants typed one passage in each of the four conditions. The passage sequence was counterbalanced across participants. Typing speeds for each of the passages was measured and averaged across participants within conditions. A repeated measures analysis of variance was used to determine whether there was a main effect of position or feedback.

In order to determine why it is that home row positioning and key strike feedback alters performance, eye behaviors, movement times and task completion times are calculated. If home row position information is important, soft keyboards without this information may have a larger number of glances that a typist directs at the keyboard. These glances will help the typist determine either whether a finger is positioned over the correct home key (the launch key) or whether the location of the key to be typed next (the target key) is in the expected position. If key strike feedback is important, soft keyboards without this information should have longer movement times where the typists do not need to glance at the keyboard. This follows since the typist will process less quickly the fact that a finger has landed on a key.

Key press and key release times will be included each time a character, number or spacebar is depressed or releases. The finger movement time between any pair of keys i and j will be derived from the key press and key release times. This time will be measured from the moment the finger leaves the launch key i until the moment that the finger arrives at the target key j. Task completion times were defined as the difference between the first key press in a passage and the last key release. Finger movement times, inter-keystroke intervals and task completion times were recorded using a program developed in JAVA 2SE. Eye movements are recorded with aid of an ASL Mobile EYE tracker.

Analyses of the finger movement times and task completions times in Experiment 2 indicated that participants were fastest when both position information and auditory feedback were included. When just finger movement times are considered, there was a significant effect of auditory feedback but not of positioning information. This was what was expected given that the speed of finger movement times is arguably largely a function of how quickly a typist perceives that a movement has been completed, something that auditory feedback, but not positioning information provides. When just the task completion times were analyzed, position information had a significant effect. The effect of auditory feedback was only marginally significant. It was expected that both factors would be significant. Perhaps the power was too small. Finally, when the eye movements were analyzed, the total scanning time was shortest when both position information and auditory feedback were available. The effects of both were statistically significant.

In summary, on the basis of the results from Experiment 1 it appeared likely that auditory feedback and positioning information accounted in part for the faster typing times of touch typists on hard keyboards as opposed to soft keyboards. In Experiment 2, this hypothesis was evaluated. Finger movement and task completion times were fastest when both auditory feedback and positioning information were present. The effect of auditory feedback appeared to impact only the finger movement times. The effect of both auditory feedback and positioning information appeared to impact the task completion times. However, the effect of auditory feedback on task completion times was only marginal. Finally, it was clear that much of the reduction in task completion times occurred because the time that the touch typists spent scanning the keyboard was smaller when both auditory feedback and positioning information was available.

It is recommended in the future that soft keyboards have both sets of feedback available, auditory (through simulated key clicks) and tactile (through home row positioning information). The gains in typing speed with these additions were models (about 10%), considered over the entire population of users the impact could be considerable.


First Advisor

Donald L. Fisher