MBS 93-19
Measuring Memory Factors in Monitoring: Reply to Kinchla
William Batchelder, David Riefer, Xiangen Hu
Kinchla has criticized Batchelder and Riefer's multinomial model for
source monitoring, primarily its high-threshold assumptions, and he advocates
an approach based on Signal Detection Theory (SDT). In this reply, we lay
out some of the considerations that led to our model, and we then raise
some specific concerns with Kinchla's critique. We point out that most
of his criticisms are drawn from contrasting the high threshold and the
Gaussian, equal-variance SDT models on yes-no recognition memory. We indicate
how source monitoring is more complicated than simple yes-no recognition,
and we questions the validity of standard ROC analyses in source monitoring.
We argue that our model is a good approximation for measuring differences
between sources on old-new detection, and that it has the advantages of
allowing goodness-of-fit tests for several submodels and the ability to
measure source discrimination, as well as detection. We also explore a
low-threshold multinomial model and show that it yields conclusions very
similarly to our original model when applied to the same data. Finally,
we examine some possibilities for applying SDT models to source monitoring
and indicate directions that seem productive for future work.