Specific effects of spatial-frequency uncertainty and different cue types on contrast detection: data and models.

If the spatial-frequency of sinusoidal signals in a contrast-detection experiment varies randomly from trial to trial, then performance is decreased compared with that in a situation where it remains constant. This spatial-frequency uncertainty effect can more or less be compensated by presenting informative cues shortly before each trial. Single-band, as well as multiple-band models, have been proposed to explain the uncertainty and cuing effects. While the latter assume that under uncertainty multiple channels are monitored simultaneously, the former propose that in each trial a single, but sometimes inappropriate, channel is selected for monitoring. Until now it is open which of these models is valid. Therefore, psychometric functions were collected under different conditions of spatial-frequency uncertainty. It appears that the size of the uncertainty effect varies with spatial-frequency. This result can be explained by a multiple-band model, as computational analysis reveals.