Expectancy of line segment orientation.

The question was asked whether briefly flashed line segments are easier to detect when presented at an expected, rather than an unexpected, orientation. Detection rates were measured in a two-interval forced choice (2IFC) paradigm that did not require the subject to identify the orientation of the line segment, only to detect its presence. The 2IFC paradigm was used to rule out bias or criterion effects. Subjects were led to expect lines in a particular or primary orientation by being presented lines with that orientation as cues before every trial, and by being tested with only that orientation during practice. Lines of the orthogonal, probe orientation replaced the primary on 25% of experimental trials. When the stimulus location was known in advance, lines of the primary orientation were detected more accurately than were probe lines, but when stimulus location was not known, detection rates were equal. Detection rates were also equal when subjects were informed of the probe at the end of the practice period, so that both orientations were expected; hence the subjects' expectations, not the probability of stimulus occurrence, are necessary for the effect to occur. Thus expecting a line of a particular orientation at a particular location facilities its detection.