Rejection Mechanisms in Recognition Memory
Lead Research Organisation:
University of Southampton
Department Name: Sch of Psychology
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Organisations
Publications
Davide Bruno (author)
(2007)
The role of metacognitive strategies in the rejection of false memories.
Higham PA
(2010)
Effects of study list composition on the word frequency effect and metacognitive attributions in recognition memory.
in Memory (Hove, England)
Higham PA
(2009)
Investigating strength and frequency effects in recognition memory using type-2 signal detection theory.
in Journal of experimental psychology. Learning, memory, and cognition
| Description | Aims and objectives of the project This research project investigated the factors contributing to the accuracy of recognition memory judgements. In a recognition memory experiment, participants first see a list of items (words, pictures) that they must remember (henceforth targets). On a subsequent recognition memory test, participants see a mixed list of targets and similar items that were not previously encountered (distracters). The task for the participant is to indicate which items were encountered before, and which are new. Theoretical interest in recognition memory has focused on the identification of targets. However, in the present research we focus on a different aspect of the recognition process, namely how people decide that an item is new. Our aim in doing this is to develop a broader understanding of the processes involved in recognition memory and to acknowledge the strategic control the participant brings to the task. The previous focus on how people decide that an item is old leads to a view that people say "new" when they cannot gather enough evidence to justify the decision that an item is old. In the present work, we sought evidence that people use active processes to determine that an item might be new. The aims of our studies were: 1) To show influences on the rates with which participants judge a distracter item to be new (measured through the false alarm rate). 2) To relate changes in false alarm rate to reported strategy use, by means of metacognitive ratings. In our original proposal, we planned a series of studies which we argued would manipulate various psychological dimensions (such as memory strength, the structure of the study list, the pre-existing familiarity of the items) which we believed would influence the strategies participants would adopt. Our expectation was that people's false alarm rates would be influenced by these manipulations, but also that participants would report using different strategies to decide whether or not an item was old or new, depending upon the context of the experiment. Main research results and significant academic achievements. Our original plan was to run 7 experiments. In the end, we ran 14 studies, using three different methodologies. Space limitations prevent a full account of these studies, and so we can only provide a summary of the studies here. Series 1 (Experiments 1a, 1b, 2, 3 in full report) were modelled on the methodology outlined in the proposal. Participants made recognition judgements about items, and also made ratings of the strategies they used in coming to their decisions. The results of these studies were mixed. Experiments 1a and 3 showed within-subject mirror effects, but the others did not. The methodology of getting participants to complete multiple ratings for each item proved overly burdensome for participants, and the ratings obtained did not reliably predict the rate of false alarm errors. Series 2 (Experiments 4, 5 and 6) were modifications of the previous methodology. As before participants made recognition judgements, but instead of independent ratings of metacognitive strategy, participants were asked to choose between a limited set of strategies for each item. Experiments 4 and 6 led to conclusive results about the role of metacognition in the rejection of distractors. In particular, using a modified rating methodology, we found that, in addition to the expected effects of study-list structure on Familiarity ratings, the effects were also manifested in ratings of Forgotten (i.e., items consistent with the structure were judged forgettable), Active Rejection (items inconsistent with the structure were actively rejected) Series 3 (Experiments 7-14) required participants to rate their confidence in the accuracy of their judgement (whether old or new), and these ratings were used to develop a signal-detection model of the accuracy of these ratings. This methodology is known as Type-2 signal detection analysis, and has not previously been successfully applied to recognition memory decisions. In particular, Experiments 12-14 (which constitute Experiments 2, 1, and 3, respectively, in Nominated Output Higham, Perfect & Bruno, in press) are the first to formally map type-1 confidence onto type-2 confidence, and to use a new methodology for examining rejection processes in recognition. In particular, using type-2 methodology, it is possible to examine metacognitive monitoring performance specifically for distractors, for which rejection processes are most common. In addition, Experiments 7, 8, and 11 (which constitute Experiments 1-3, respectively, in Nominated Output Bruno, Higham, & Perfect, under review) we investigated the conditions under which lures are differentially rejected as a function of repetition. Across the studies, we found a differential rejection only under conditions of poor overall subjective memorability. We suggest that when subjective memorability is low, people switch their metacognitive strategy on an item-by-item basis, and that this generates the observed rejection patterns. Dissemination activities Our strategy has been to target high impact outlets in the field. We have one paper accepted in Journal of Experimental Psychology: Learning, Memory and Cognition, a second paper under revision having been submitted to Memory and Cognition, and a third paper in preparation. The work has also been presented at international conferences (three papers to the Psychonomic Society Meeting in Long Beach and Chicago, and another to the British Psychological Society meeting in Aberdeen) and at universities in the UK and US. |
| Exploitation Route | The aim of the work conducted in this project was to inform basic psychological theorising in the area of human memory. We believe that there are a number of unique strengths of the work conducted. We have demonstrated a reliable within-subject mirror effect, and developed a theoretical account of why previous attempts to find such an effect have proven elusive. We have also developed a new methodological approach to understanding recognition memory - Type 2 signal-detection analysis - and have shown its efficacy in theory development. We believe that both these findings will have significant impact upon theoretical work in recognition memory in the years to come. |
| Sectors | Other |