The Measurement of Tapping Performance on Mobile Devices Using Various Gestures under Different Postures

• Main Authors: Bo-Tsun Lin, 林柏村
• Other Authors: Jui-Feng Lin
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/g6kqf5

碩士 === 元智大學 === 工業工程與管理學系 === 103 === To increase user satisfaction and tapping performance while using touchscreen cell phones. Developers have been devoted to study the tapping performance when users interacting with a cell phone. Several studies have assessed and given suggestions for designing appropriate button sizes. However, due the limitation of method used, these studies had difficulty to reveal the individual performance of movement speed and accuracy. To overcome the limitation, Aguilera (2015) and薛沂禾 (2013) applied the ballistic movement method to measure tapping performance in terms of movement time and accuracy under different conditions with varied body postures, hand gestures and mobile devices with different sizes. Although, due to the unavalable information of movement distance, two study could not validate the application of the ballistic movement time and variability models, their results showed that tapping movement time and tapping error are differnet in different environment. In handedness finger tapping, the movement time and variable errors were greater in walking situation than those measured when sitting on train (Aguilera, 2015). In sitting situation, the movement time and variable errors were greater in handedness thumb tapping than those measured when handedness finger tapping (薛沂禾, 2013). To provide movement distance information and measured tapping conditions that were not measured, this study aimed at using a motion capture systems and the ballistic movement method to measure tapping performances in terms of moving time and accuracy under different conditions of body postures, hand gestures and mobile devices with different sizes and to validate the application of the two ballistic movement models. This study was divided into a pilot study and a formal study. The pilot study aimed at testing the initial experimental design. In the formal study, twelve participants were recruited to perform ballistic movements on mobile devices. A motion capture system (OptiTrack Flex 3) was applied to record the trajectory of finger movement while tapping. Independent variables were two postures, six kinds of gestures to handle devices, and four sizes of devices. Dependent variables were the ballistic movement time, constant errors and variable errors both measured in horizontal and vertical direction. Several results reported by this study. First of all, the ballistic movement time varied according to postures, gestures and sizes. The ballistic movement time was less when tapping at the starting point area and when standing. It was greatest when tapping 10.1 inch tablet and shortest when tapping 4.7 inch cell phone. Furthermore, it was greater when tapping at the upper left and lower right areas of the screen in a portrait mode, and when tapping at the upper center and lower center of the screen in a landscape model. Second, the endpoint variable errors varied according to postures, gestures and sizes as well. The variable errors were less when tapping at the starting point area. It was greatest when tapping 10.1 inch tablet and shortest when tapping 4.7 inch cell phone. Furthermore, using handedness finger and thumb variable error of horizontal direction was greater when tapping at left and right edge of the screen in a portrait mode, and using two thumbs when tapping at center of the screen in a landscape model. When using handedness finger and thumb, the variable error of vertical direction was greater when tapping at upper and lower of the screen in a portrait mode, and when tapping at the left lower and right lower of the screen in a landscape model. Third, the constant error was greater when tapping at the edges of touchscreen, and the direction tendency to toward center of the screen. Finally, although the ballistic moving time model could describe the linear relationship between the moving time and root mean square of distance, due to low variation of moving distance and other non-control variables (like moving directions) the ballistic movement variable model had lower prediction of the data. This study with studies of Aguilera (2015) and薛沂禾 (2013) showed that the movement time was greatest in walking situation and least in standing posture. The movement time and variable errors were greater in handedness thumb gesture than handedness finger. When tapping with two thumbs, movement time and variable errors were least toward starting points, the touchscreen center, upper and lower area. Movement time and variable errors increased with increased size of mobile devices. These results could be applied when designing appropriate buttons in various conditions. Keywords: Smart mobile device, Ballistic movement method, Touchscreen, Aiming movement, Tapping performance.