The influence of emotion on remembering and forgetting

How emotion influences the ability to control what we remember and forget remains unclear. The objective for the work described in this thesis was to develop a better understanding of the links between emotion, memory and memory control. This was achieved by acquiring behavioral and ERP data in a series of directed forgetting (DF) experiments. The DF item-method was used throughout, with different retrieval task requirements. These different retrieval requirements permitted an analysis of how emotion links to memorability, to directed forgetting, and to two different retrieval processes – recollection and familiarity. A key goal in this work was to exert tight control over stimulus parameters and maintain that control consistently over studies. The intention was to permit contrasts between outcomes across the set of experiments with fewer degrees of freedom than is the case when comparing outcomes among published experiments from different researchers. This is an important consideration because the relevant literature contains many contradictory findings for which there are several competing explanations. Alongside several parameters over which control is commonly deployed, in the experiments described here semantic relatedness was also controlled for. This has not been done consistently in studies of the links between emotion and memory, and the outcomes here and elsewhere suggest that controlling for this factor, which likely co-varies with emotion, is important. Key behavioral findings in these experiments were: mixed evidence for superior memory for emotional materials (with this term being used here to refer to positive and negative valence words); no evidence that DF varies according to valence, a consistent liberal response criterion for emotional materials, with the exception of Experiment 1 and when valence is manipulated at study only (Experiment 5); and mixed findings for the influence of emotion on recollection and familiarity. In addition, and for the first time, the data in the final experiment suggest that the liberal response criterion associated with emotional materials can be attributed to processes that occur at the time of retrieval rather than at the time of encoding. In three event-related potential (ERP) experiments the test phase data provided no new insights over and above what can be inferred from the behavioral data described above. There were, however, informative outcomes from study phase data where ERPs were acquired timelocked to the cues to remember/forget that followed the critical items. Neural activity elicited by remember and forget cues varied according to emotion, implicating valence in the strategies people use to remember and forget material. In summary, across a series of experiments there was no behavioral evidence that memory control (operationalized as control over remembering and forgetting) varied with valence, but the ERP data indicate that remembering and forgetting does vary with emotion. Moreover, in what is perhaps the other substantive new insight in this thesis, the data in the final experiment suggest that the liberal response criterion associated with emotional materials can be attributed to processes that occur at the time of retrieval rather than at the time of encoding