Context effects are not explained by bottom-up theories of accounting. Irving Biederman performed experiments that demonstrated dramatic context effects. For example, Stephen Palmer carried out an experiment in which the participants were asked to identify objects after they were shown either a relevant or irrelevant context. They might be shown a scene of a baseball game, followed by images of a baseball, car, and a phone. The stimuli that was most relevant to the context, the baseball, was recognized quicker than those that were irrelevant, car and phone.
Perceptual constancy gives evidence that high-level constructive processes occur during perception. As lighting conditions change, the color of objects, such as bananas or cherries, appear to remain constant. Even when there is not enough light to even stimulate cone cells and give the sensation of color, bananas and cherries are still perceived as yellow and red, respectively.
Configural-superiority effect is another compelling context effect. This effect is demonstrated by the decrease in response time by participants when identifying objects in complex configurations over objects in isolation. For example, four diagonal lines are shown and participants are asked the location of the odd line indicated by its different orientation. In another condition, an "L" shaped fixed context is added; this creates three triangles and another object made of three lines. Participants were able to discern the odd three line object from the set of triangles quicker than the single diagonal lines alone.