Home Leadership Tasks Publications Biomarkers
Gain Control: The Contrast-Contrast Effect Task (CCE)

 

Gain control has been successfully studied using the Contrast-Contrast Effect (CCE) task in which contrast sensitivity for a ringed target can be influenced by the contrast of a circular surround. Participants are asked to match a variable contrast patch to a central patch. When the surround is high-contrast, the inner target is perceived to be of lower contrast than when the same target is perceived without a surround. This task is ideal for examining gain control in schizophrenia because: 1) reduced gain control, or contextual modulation, would be indicated by more accurate contrast judgments regarding the inner circle compared to controls; and 2) there is already evidence for reduced spatial context effects in vision in schizophrenia. The following is a description of the version of the task available below.

On each trial, subjects view a circular region (1.3° diameter) of blob-like shapes (8 c/deg. bandpass-filtered noise) presented for 1000 ms at an intermediate contrast level (40%). After a 200-ms blank period (modified task), an isolated central comparison patch was presented, and subjects indicated whether the first or the second patch had higher contrast. The contrast of the comparison patch is varied to provide a psychometric function for the contrast perception. The first patch is presented either in isolation or surrounded by a high contrast (95%) annulus of the blob-like shapes (8° diameter).

Participants completed two blocks of trials: one with the first patch presented in isolation and one with the first patch surrounded by a high contrast stimulus. Participants should complete the "no-surround" block before the surround block. Within each of the two blocks, 180 trials are presented, drawn psuedorandomly from one of 5 "streams," where the information used to compute trial types in each stream was kept separate.

  • Two streams (N=40 trials each) in which a staircase procedure is used to move participants to 50% accuracy. In both streams, the first trial comparison patch is presented at a 30% contrast difference from the target patch (which is always 40%). In one stream ("S2") the first trial comparison patch is 30% higher than 40% (e.g., 70%). In the other stream ("S3), the first trial comparison patch is 30% lower than 40% (e.g., 10%). If the participants respond correctly, then the next comparison patch is more similar to the target patch. If they respond incorrectly, the next comparison patch is less similar to the target patch. For the first 15 trials, the contrast is changed in increments of 5%, for the next 15 trials, 2.5% and for the last 10 trials the increment is 1%.
  • Two streams (N=40 trials each) in which a staircase procedure was used to move participants to 75% accuracy. In both streams, the first trial comparison patch is presented at a 30% contrast difference from the target patch (which is always 40%). In one stream ("S4") the first trial comparison patch is 30% higher than 40% (e.g., 70%). In the other stream ("S5"), the first trial comparison patch is 30% lower than 40% (e.g., 10%). If the participants responded correctly 3 times (even with interleaved errors), then the next comparison patch is more similar to the target patch. If they respond incorrectly, the next comparison patch is less similar to the target patch. For the first 15 trials, the contrast is changed in increments of 5%, for the next 15 trials, 2.5% and for the last 10 trials the increment is 1%.
  • One stream of "catch" trials (N=20 trials) in which the comparison patch is either 70% contrast of 10% contrast. These trials are included to assess the ongoing attention level of the participants, as the high contrast difference from the target patch should make them easy for participants who are attending to the task.

Scoring: The data from each subject can be scored in several ways. First, the average contrast for the last 10 trials in both of the streams can be compared (ignoring catch trials) for the no-surround and surround condition. If gain control mechanisms are working properly then, the contrast values for the last 10 trials in the no-surround condition should be closer to 40 than in the surround condition. Second, the average contrast for the last six reversals in both streams can be compared (ignoring catch trials), with a similar interpretation. Lastly, one can fit with a cumulative Gaussian function to estimate the accuracy (bias/intercept) and precision (slope) of the subject's contrast perception, and this can be done separately for trials with and without the high-contrast surround region and separately for each version of the task. The bias should be shifted for the surround versus the no-surround condition. Currently, the outputs of the tasks contain the trial by trial data, and must be analyzed off-line to compute the summary statistics described above. We will be modifying the scripts to include summary statistics (e.g., average contrast in last 10 trials or last 6 reversals) at the end of the task.

These versions are currently being used in our test-retest reliability, and could change based on the results of that study. In particular, it is likely that the task will be shortened if the reliability results support such a change.