Monitoring recognition memory: A signal detection analysis of internally and externally generated influences on discriminability and response bias
This dissertation contributes to a growing body of research that attempts to bridge the chasm between basic and applied memory research. Its basic approach is to use signal detection theory to analyze higher-level cognitive components that influence recognition memory. The work consists of three research papers that examine the effects of internal and external sources of memory on discriminability and response bias in order to better understand memory in the real world. Paper one provides new evidence supporting a basic assumption of reality-monitoring theory that certain cognitive operations are important for knowing whether or not a memory comes from an internal source. This research demonstrates that people are more susceptible to false memories after completing mindfulness training because their reality-monitoring accuracy is reduced. Paper two examines the verbal overshadowing effect (where people are worse at correctly identifying someone from a police lineup after providing a verbal description of a face) with receiver operating characteristic (ROC) analysis to determine if that well-known effect is due to differences in actual memorability rather than differences in response bias. This research indicates that internally- generated information can be confused with an external memory source when the internally-generated information is not sufficiently detailed. Paper three examines the cross-race effect wherein memory is worse for a person of a different race than a person of the same race. This research indicates that although memory is worse in terms of discriminability, high-confidence identifications are just as reliable for a cross-race face as for a same-race face.