Institute for Mathematical Behavioral Sciences

Recognition memory experiments are an important source of empirical constraints for theories and models of memory. Unfortunately, standard methods for analyzing recognition memory data have problems that are often severe enough to prevent clear answers being obtained. A key example is whether longer lists lead to poorer recognition performance. The presence or absence of such a list length effect is critical test of competing item- and context-based theories of interference, but remains an unresolved empirical issue, largely because of the weaknesses of the standard analysis. In this paper, we develop a new Bayesian method of analysis that overcomes the problems. We report data from a new recognition memory experiment that manipulates list length, as well as the better understood manipulation of word frequency, and present both standard and Bayesian analyses of the data. The comparison of the two methods allows us to highlight the advantages of the Bayesian approach in inferring the values of psychologically meaningful variables, and in choosing between different models representing different theoretical assumptions about memory.