Exploring the Optimal Number of Trials and Participants for Touchpad Testing Using Fitts-based Tasks

Abstract

With touchpads being the primary input device for many laptop users, the importance of ensuring that laptops have the most functional touchpad has continued to grow. Human Factors practitioners are often tasked with ensuring that touchpads are accurate and efficient input devices. Tasks based on Fitts law to assess point and click performance are commonly used for testing touchpads, however there does not seem to be a widely used standard regarding the number of trials and participants. Testing with more trials and participants than necessary can be time consuming and costly for industry practitioners. This research explored varied numbers of participants and Fitts task trials with the goal of determining the optimal practices to maximize efficiency. Preliminary results indicate that when testing performance with a multidirectional Fitts test, twelve participants is acceptable for device comparison with at least 48 trials. For a sample size of 24 participants, there is evidence to support reducing the number of trials to 24. Implications for examining other testing scenarios for optimal sample size and trial number with the prescribed methodology are discussed.