REMI and ROUSE: Quantitative Models for Long-Term and Short-Term Priming in Perceptual Identification

摘要: textabstractThe REM model originally developed for recognition memory (Shiffrin & Steyvers, 1997) has recently been extended to implicit phenomena observed during threshold identification of words. We discuss two models based on Bayesian principles: a long-term priming (REMI; Schooler, Shiffrin, Raaijmakers, 1999), and short-term (ROUSE; Huber, Lyle, Ruys, in press). Although the tasks are same, basis differs models. In both paradigms we ask whether prior study merely reflects bias interpret ambiguous information certain manner, or instead leads more efficient encoding. The observation ‘both-primed benefit’ two-alternative forced-choice appears show that processes present. However, REMI illustrates both-primed benefit is not necessarily indicative an increase perceptual sensitivity but might be generated by criterion bias. ROUSE demonstrates how amount attention paid prime, consequent effect upon decision making, may lead reversal normal is conditions.