Using GOMS to predict the usability of user interfaces of small off-the-shelf software products

The design of user interfaces and how usable they are, are both important research topics in computer science. This thesis is a research effort aimed at exploring the whole concept of usability and measuring the quality of a user interface in terms of how usable it is. Usability means how easy a system can be learned and used. In order to have usable products, they must be initially designed with usability in mind. A survey of methods for designing user interfaces which incorporate usability are outlined and they include some or all of the principles for designing for usability, proposed by various authors. Evaluating the quality of existing interfaces can be done by various methods.The method used in this dissertation is the GOMS (goals, operators, methods and selection rules) approach. This model was initially proposed by [Card, Moran & Newell 83] and the approach is based on constructing an explicit model of the user's procedural knowledge, entailed by a particular system design. [Kieras & Poison 85] expanded this model to suggest that quantitive measures defined on this explicit representation of the user's knowledge can predict important aspects of usability. The predictions are obtained from a computer simulation model of the user's procedural knowledge that can actually execute the same tasks as the user. To test the reliability and accuracy of the GOMS model predictions, the author carried out a pseudo-experiment on four inexperienced users using two different types of word-processors. The actual results from the experiment were compared with the GOMS predictions.The GOMS model was found to have some limitations and some enhancements to the approach are proposed. It was also found that the experiment had some limitations and improvements for a better experiment are proposed