Touch-level model (TLM): evolving KLM-GOMS for touchscreen and mobile devices

In the early days of personal computing, a model called "Goals, Operators, Methods, and Selection Rules" (GOMS) provided a method with which to quantify user interaction with a system. [2] Soon after, the Keystroke Level Model (KLM) allowed decomposition of complex tasks into atomic, benchmarked 'operators' that could generalize and estimate interactions on any keyboard interface. However, this the model is dated and unsuitable for today's modern devices. These touchscreen interfaces are now pervasive [4], but the HCI implications of this shift are not fully researched; an imperative exists for research in human-computer interaction to study these interfaces and the interactions of the people using them. [3] In this paper, we introduce new operators and other modifications to KLM-GOMS to accommodate modern touchscreen interfaces. We call these additions, together with updates to the existing KLM operators, the Touch Level Model (TLM). We propose that this model can be employed to model human task performance on a constrained-input touchscreen device and, with proper benchmarking, accurately predict actual user performance. Our goal is to provide a means for quantitative analysis of touchscreen interfaces for HCI research and to enable description of interactions with such interfaces in terms of these operators, thus predicting user performance with less need to create prototypes or perform user studies.