In this project we will investigate our ability to assess the recency of occurrence of stimuli. We will use a simple immediate memory probe task: on each trial a list of between one and four stimuli (eg faces) is shown to a participant, followed by a short pause and a test (probe) stimulus. The participant makes a rapid "yes"/"no" decision as to whether the probe was one of the list items. Our previous work has shown that if the item was not in the list, but has been seen recently (eg on the last two trials), then the 'no' response is slower and more errors are made than if the probe item has not been seen for a number of trials: this difference provides us with an index of apparent recency. We will control the stimuli encountered by our participants between "study" (when they last saw stimulus X) and "test" (when X is presented as a "no" probe on the current trial). Our model predicts that interpolated items will influence the estimated recency of the probe depending on the similarity relationship between them, familiarity of the stimulus set, and the associative linkages created by earlier encounters with other items.

Individual data sets collected via laboratory run experiments. We ran behavioural experiments on normal volunteers to explore the memory mechanism that enables us to assess how recently we have seen an object (e.g. a person, a pattern, a word). We used an immediate probe recognition paradigm: on each trial a sequence of two to five objects was presented (fast enough to discourage rehearsal), followed by a mask, followed by a probe object. The participant decided whether the probe was one of the objects in the sequence just presented, or not (in which case it had occurred in an earlier sequence).