Spatial and Temporal Factors Affecting Human Visual Recognition Memory

The current thesis investigated the effects of a variety of spatial and temporal factors on visual recognition memory in human adults. Continuous recognition experiments investigated the effect of lag (the number of items intervening between study and test) on recognition of a variety of stimulus sets (common objects, face-like stimuli, fractals, trigrams), and determined that recognition of common objects was superior to that of other stimulus types. This advantage was largely eradicated when common objects of only one class (birds) were tested. Continuous recognition confounds the number of intervening items with the time elapsed between study and test presentations of stimuli. These factors were separated in an experiment comparing recognition performance at different rates of presentation. D-prime scores were affected solely by the number of intervening items, suggesting an interference-based explanation for the effect of lag. The role of interference was investigated further in a subsequent experiment examining the effect of interitem similarity on recognition. A higher level of global similarity amongst stimuli was associated with a lower sensitivity of recognition. Spatial separation between study and test was studied using same/different recognition of face-like stimuli, and spatial shifts between study and test locations. An initial study found a recognition advantage for stimuli that were studied and tested in the same peripheral location. However, the introduction of eye-tracking apparatus to verify fixation resulted in the eradication of this effect, suggesting that it was an artefact of uncontrolled fixation. Translation of both face-like and fractal stimuli between areas of different eccentricity, with different spatial acuities, did decrease recognition sensitivity, suggesting a partial positional specificity of visual memory. These phenomena were unaffected by 180 degree rotation. When interfering stimuli were introduced between study and test trials, translation invariance at a constant eccentricity broke down