Discriminability Task

This task consist of a number of trials in which a pair of stimuli is shown briefly, after which the participant responds with a key press to judge whether the stimuli shown were the same or different.

Timing of a trial

How long each individual trial takes depends on the following parameters, in chronological order. They are each specified in milliseconds. Note that the actual duration on the participant's screen depends on their monitor refresh rate, among other things. Some can be set to 0 to negate them. stimulus durationThis is how long the stimuli themselves are presented blank durationThis is a period during which the screen will be empty right after the stimulus mask durationThis is how long each pair of masks is shows (see below) pause durationThis is the break between the end of the last trial and the beginning of the next

Sampling of the stimulus pairs

The pairs of stimuli shown during the task are sampled from the stimuli chosen in the Stimuli tab of the task. If the total number of possible pair combinations of the stimuli is less than the maximum number of trials parameter, all combinations will be used, if it is larger, only some (random) pairs will be shown. The proportion of identical pairs parameter, if larger than 0, creates extra trials in which the same stimulus is shown on both sides. This is important for the same/different response to be meaningful.

Masking

In order to make the perception of the stimuli more challenging, it is possible to display a series of masks covering the space taken up by the stimuli. These masks will be auto-generated from the stimuli by dividing them in squares and scrambling these parts. In order to turn off masks, simply set the number of masks parameter to 0. The mask spatial frequency parameter determines how many squares the stimuli should be partitioned into. Note that the stimulus resolution has to be divisable by this number, or the task won't work. For example, for a picture stimulus of 200 by 200 pixels, with a mask spatial frequency of 8, the masks generated will consist of 64 squares, each 25 by 25 pixels.