Title:  DEVELOPING A MEMORY AND DECISION MODEL OF EYEWITNESS IDENTIFICATION ABSTRACT:  Clark (2003) has recently developed a memory and decision model that generates quantitative predictions for response probabilities in eyewitness identification tasks.  The model assumes that a witness, when presented with a lineup, compares each lineup alternative with his or her memory for the perpetrator of the crime.  Decisions as to whom to pick, or whether to pick at all, are based on a weighted combination of the value of the best match to memory and the degree to which the best match is better than the other alternatives.  Although this model is certainly an oversimplification of the identification task, it has shown itself to be a useful tool for generating predictions, guiding research, and asking new questions about how witnesses make identification decisions, and why they make mistakes.  New variants of the model, which vary the assumptions about the representation of information, the computation of matches, and the decision processes, are currently being developed.