Vividness & accuracy

Abstract

Rich or vivid experiences are widely believed to be better remembered but the relationship between vividness and memory is unclear. We displayed photographs of natural scenes in two experiments utilising a previously reported continuous memory task1 to examine the relationships between vividness, memory, and confidence in memory judgments. Experiments were conducted using E-Prime (version 2.0 Professional, Psychology Tools Inc.), on a laptop PC. Photographs were presented in landscape orientation on a white background at a resolution of 640 × 425 pixels. Each trial was made up of a cross location paired with a natural scene photograph. Photographs were rated for vividness either at encoding (when first experienced; Experiments 1 & 3) or at retrieval (when remembered Experiments 2 & 4). Presenting the related image cues retrieval of the position of the cross on the circle only if the participant recollects related contextual information; memory accuracy is measured as the degrees of error away from the actual location. Location errors within ± 90° of the target are used to calculate the proportion of trials associated with recollection, i.e., the recollection rate. thus providing an assessment of the quality of retrieval. The recollection threshold, expressed in degrees from the target location, indicates the exact point at which recollection begins to contribute to responding and is calculated using linear regression to determine when the slope of the distribution of errors changes from zero (i.e., is no longer flat because responding is not solely reliant on guessing); the scale parameter of the Cauchy distribution represents indicates the shape of the location error distribution and provides a direct measure of memory precision. Experiments 1 & 2 used 120 photographs (optimised for colour and contrast) and consisted of ten cross location-natural scene relational encoding blocks (each containing 12 trials, pseudo-randomised for subject type) and ten cross location-natural scene recollection test blocks with confidence ratings and were presented such that a single study block was followed immediately by the corresponding test block, each participant completing 120 trials Photograph presentation order within each trial block during both study and test phases, and trial block presentation order within each experiment, were chosen at random. In Experiments 3 & 4 we used a 240-image set of photographs, repeating Experiment 1 with the addition of 120 natural scene photographs that were less visually salient, being partially desaturated for colour and not optimised for contrast (sharpness); in this case the experimental paradigm consisted of 15 cross location-natural scene relational encoding blocks (each containing 16 trials, pseudo-randomised for both subject and colour saturation) and 15 cross location-natural scene recollection test blocks with confidence ratings, meaning that each participant completed 240 trials.