Search

In Eqs. (1-2), t is the pointing time, A is the amplitude (distance) to the target, W is the target size and a, b are constants that depend on experiment conditions. ID is larger as A is larger and/or W is smaller. Values of a and b in Eq. (1) are determined by sampling (A, W, t) data and applying the linear regression analysis to the data. Eq. (2) shows that ID values are the same for (A, W) and (nA, nW) where n > 0. This research is motivated by recent smart phones that employ touch UIs. Compared with other touch screen devices such as tablet PCs, mobile phones have smaller screens so that widgets on mobile phone screens are likely to be smaller. Widgets can be designed for devices with various screen sizes so that theoretical ID values in Eq. (2) are consistent among the devices: larger/smaller sizes & distances for larger/smaller screens. If ID in Eq. (2) is an appropriate index of actual pointing difficulty independently of screen sizes, users’ pointing performances on the same device are consistent among widget designs (A, W) and (nA, nW): note that a, b in Eq. (1) are constant (independent to ID) so that a, b must be the same for two data sets sampled with the two widget designs (A, W) and (nA, nW). The aim of this research is to investigate whether the above is true: appropriateness of the ID formulation in Eq. (2) is evaluated from the viewpoint of dependency on screen sizes, by experiments with participants. Limitations of Fitts’ law have been researched and extensions have been proposed. For example, an extension for 2D pointing tasks was proposed (MacKenzie & Buxton, 1992). Our research aims at investigating possible limitations on screen sizes. A related research was previously reported (Oehl et al., 2007). They investigated how display size influenced pointing performances on a touch UI and reported that in large displays a fast and comparably accurate execution was chosen in contrast to a very inaccurate and timeconsuming style in small displays. In their research the size of small screen was 6.5”, and only a large screen touch UI device was utilized for user experiments: screen sizes were