Models and Theories in Human-Computer Interaction/GOMS may be useful, but is it practical?

GOMS may be useful, but is it practical? (Cindy Marinak)

HCI Models, Theories, and Frameworks define GOMS (Goals, Operators, Methods, and Selection rules) as a predictive model used to analyze error-free task efficiency for expert users. The model describes a user’s tasks by: the user’s goals to accomplish the task; the available operators (actions); the methods, comprised of subgoals and operator sequences; selection rules if there are multiple methods to accomplish the same goal. (Carroll, 2003) While I understand the value of collecting performance metrics, GOMS analysis does not appear to be a common approach used by software project teams.

The variations of GOMS analysis do not seem practical for use in many software companies. In my experience, small to mid-sized software organizations often hire generalists. These practitioners possess diverse skills such as user experience (UX) design, high-level user interface (UI) development capabilities, and broad versus in-depth knowledge of research techniques. Generalists often have some experience with common research methods such as heuristic evaluations and usability tests; some may not be aware of GOMS and likely do not possess experience conducting the analysis and interpreting the results. Are practitioners in the human-computer interaction (HCI) field familiar enough with GOMS models to use the technique for project research? If there is not a researcher on staff to conduct GOMS analysis, is there a need or willingness to hire an expert for a particular study?

In my experience, incorporating research into the software development lifecycle (SDLC) is challenging. The research method(s) proposed for a given project are based on factors such as the project release cycle, product maturity, budget, available resources, and goals of the study. A benefit of GOMS analysis is that it does not require end user participation. A limitation is that the analysis focuses on efficiency of skilled users’ error-free task completion. (Carroll, 2003) Efficiency is only one aspect of usability. Research without considering novice and intermediate users seems contradictory to many organizations striving for usable interfaces for all users.

Perhaps awareness of the GOMS model; practitioners’ exposure to tools that facilitate performing GOMS analysis; educating stakeholders about the benefits of the model, will promote an industry trend toward incorporating GOMS into product research capabilities.

Work Cited

Carroll, J. M. (2003). HCI Models, Theories, and Frameworks. San Francisco: Morgan Kaufmann Publishers.