The effect of practice on the visual detection of near-threshold lines

Two observers practised to detect small target lines of varying luminance presented either within a context of collinear inducing stimuli, or without the context in separate blocks. A two-alternative spatial-forced-choice procedure using the method of constant stimuli was employed. For blocks of 500 trials, reflecting individual performance on five successive days of training, the percentage of correct responses, and the response times were analyzed. After several thousands of trials, i.e. several weeks of practice, both observers managed to detect targets presented at their strongest luminance within the context condition. Without the context, these targets remained undetected. Response times (RT) vary nonsystematically during training. Once detection is observed at the highest target luminance, the accuracy of the individual responses (percentage of correct responses) systematically increases, and processing speed (RT) systematically decreases with increasing target intensity. These results show that, within the appropriate perceptual context, practice can lead to a better detectability of visual stimuli presented at luminance levels near detection threshold. This improvement is reflected by an optimization of the visual integration of the different luminance levels of the target (sensory coding), and processing speed (response routine) during training.