State-Dependent Learning and Retrieval: Methodological Cautions and Theoretical Considerations
John Kihlstrom, Frederick Evans · Psychology Press eBooks · 2014
In many experiments with drugs a subject learns new information in one of two “ states” : when the drug in question is given and is exerting its effect on the central nervous system, or when a placebo is given and the drug is not exerting its effect. In the same way, retention may be tested in either the drug or placebo state. In such a paradigm it is important to consider the possibility of state-dependent learning (SDL)— that is, whether retention is better when the subject is tested in the same state as that in which learning originally occurred. Table 9.1 shows the 2 x 2 factorial design that is standard in the SDL literature and describes the outcomes that may be observed. If SDL occurs, retention in the d rug -d rug (DD) combination should be superior to that in the placebo-drug (PD) combination, and retention in the placebo-placebo combination (PP) should be superior to that in the drug-placebo combination (DP). If only SDL occurs, the two control conditions (PP and DD) should be equivalent. However, invariably the drug treatment has some other effect as well, and it is important not to confuse these effects with SDL. A drug may either impair or facilitate initial learning or subsequent retention, which would be reflected in the analysis of variance as a significant main effect of either learning or retention state (separately or in combination); evidence for SDL emerges only when the learning and retention states interact.