Thursday Oct 8, noon-1pm, 2203 Dunton Tower, Carleton University

Title: Modeling novice to expert performance with a modeling compiler
 Frank E. Ritter (frank.ritter@psu.edu)
 College of IST, Penn State

Abstract:In this talk I'll present a high level behavior representation language (called Herbal) that helps model performance on an spreadsheet task.  The task takes novices about 25 min. to do and experts about 18.  We created a tool to create models from a hierarchical task analysis.  The tool creates 12 models, a completely noivce model that has to proceduralise everything, and then 11 levels of initial expertise, ranging from basically a novice to a someone that can do the ask using just procedures.  The models predict the novice to expert transition, and the models got created particularly quickly.  The models fit the data fairly well (surprisingly well), and the talk suggests that we can now model novices and their learning using a GOMS-like model, and that behavioral modeling languages should be used more.  If there is time, I'll also briefly mention our tutors, including one for rifles that we have tested and ones we are working on for health care and maintenance.

Biography
Dr.  Ritter is a professor of IST, of Psychology, and of CS and Engineering at Penn State, and an associate investigator at NSMRL. He researches the development, application, and methodology of cognitive models, particularly as applied to interface design, predicting the effect of behavioral moderators, and understanding learning.  He was a staff scientist at BBN before going to CMU for his PhD. He is currently the Penn State PI for an ONR project looking at tutoring.  He is also interested in usability of modeling in general, and has helped develop the CoJACK architecture as an exploration in other ways to create models quickly.    His books include one on applying cognitive models in synthetic environments, published by the Human Systems Information Analysis Center (HSIAC, WPAFB, 2003), and a book on order effects on learning (2007, Oxford). He contributed to a National Research Council report on how to use cognitive models to improve human-system design (Pew & Mavor, eds., 2007), and he has recently published two textbooks on what psychology do systems designers (and modelers) need to know (Springer), and one on the practical aspects of running studies (Sage).  He currently edits the Oxford series on cognitive models and architectures for Oxford University Press, and is an associate editor of Human Factors and IEEE Human-Machine Systems.