Abstract
Four modelers present tools based on or for cognitive human performance modeling. Myers introduces a new statistical technique for testing the similarity of sequential behavior across conditions. This technique promises to solve what Anderson (2002) regarded as the non-determinism problem of modeling behavior at the 100–ms level of analysis. John presents a programming-by-demonstration system that creates keystroke level GOMS models in ACT-R. Her approach enables those not trained in cognitive science to build predictive models of human performance. Salvucci's work expands on John's system by applying predictive modeling techniques to in-vehicle devices. His work integrates models of device use with a rigorous model of driver behavior to predict driver distraction and performance. Finally, Gray introduces Cognitive Metrics Profiling (CMP) — a model-based approach that produces theory-based estimates of cognitive workload. CMP holds the promise of predicting transient changes in cognitive workload that occur in a dynamic task environment.
Get full access to this article
View all access options for this article.