A methodological approach to study the role of memory and attention on visual perception

Computer-based experiments in visual perception studies open up great opportunities for using natural and processed images and sophisticated experimental techniques. Meanwhile, the apparent simplicity imposes additional obligations on the correct interpretation of results. We investigated popular psychophysical methods used to study the role of memory and attention in visual perception. Memory mechanisms were studied by comparing sequential and simultaneous presentation of the same stimuli in a curvature estimation task of real and interpolated lines. In contrast to sequential presentation, simultaneous presentation revealed the illusion of straightening of interpolated images and lower curvature discrimination thresholds. The role of attention was examined by comparing the performance in single and dual tasks: the subjects were asked to recognize the shape of tests and distractors located at different distances from the tests. The stimuli had low contrast and a short presentation time. We assumed that performing the double task would affect the distribution of attention and worsen test recognition. However, the double task produced no significant deterioration in the recognition of the test compared to the single task, except for the number of nonrandom errors that showed a significant decrease. The distances where the deterioration was detected greatly exceeded the generally accepted ones obtained in similar experiments using other observation conditions. To conclude, the influence of attention in both tasks was related to stronger deterioration in recognition ability when distractors were similar in shape to the tests but different in orientation. Thus, the application of different methods may reveal discrepancies in the results.