Precis of project Ernestine or an overview of a validation of GOMS

Abstract

A major test of the applicability of GOMS (goals, operators, methods, and selection rules) to real world design problems is summarized. Project Ernestine compared the worktimes of telephone company operators on two different workstations, and tested the validity of GOMS models for predicting and explaining real world performance. The authors provide an overview of the methodology of the study, the empirical data, and the GOMS models. The GOMS models predicted performance with exceptional accuracy. The data provided three results. The new workstation was slower than the old one. The difference was not constant but varied with the type of call. There was no evidence of learning in data spanning four months and 78,240 calls. The GOMS models predicted the first two results, and explained all three.