Factors Affecting the Likelihood of Constrained Retrieval

As a kind of controlled cognition, strategically reinstating an encoding process to facilitate retrieval of specific subsets of information is more cognitively costly than more automatic forms of retrieval. Although strategic reinstatement of encoding operations is cognitively costly, it may offer no memorial advantage when familiarity-based retrieval is diagnostic or when following events that fail to promote distinctive encoding. In three experiments, I investigated conditions that vary the likelihood and benefit of encoding context reinstatement using the memory-for-foils paradigm (Jacoby, Shimizu, Daniels, & Rhodes, 2005a). In Experiments 1-2, I tested whether the likelihood of constrained retrieval varies as a function of the diagnosticity of familiarity by varying word frequency (Experiment 1) and the need for recollective details in the context of an associative memory task (Experiment 2). In Experiment 3, I tested whether the likelihood of constrained retrieval varies as a function of cue distinctiveness. Throughout, I also evaluated whether constrained retrieval actually benefits memory. Overall, I created variations in the likelihood of constrained retrieval and found evidence to support the view that constrained retrieval benefits recognition memory, particularly in the case of memory for deeply processed items.