[Nonlinearity of interference in recognition memory for multiple contexts].

It is logically inferred that in a recognition judgment, subjects unconsciously discriminate an experimental episode from past episodes concerning the test item. This implies that the interference effect reported in the list specific recognition task (Anderson & Bower, 1974) must also appear in the standard recognition judgment. This paper investigated the nature of such an interference effect. In the experiment subjects were presented with a target list following non-target lists, and then performed a YES-NO recognition test. Each previously-studied list contained both targets and distracters in a different order. The subjects were divided into four groups according to the number of prior study lists (0, 1, 2, 3). Results indicate that recognition performance deteriorated as the number of prior lists increased. The most important result was that the rate of increase of the interference effect, especially when estimated based on the increment of false alarms, decreased as the number of prior lists increased. This paper calls this phenomenon non-linearity of interference effect.