Implicit learning of spatial context in visual search

Abstract

In the present experiments the paradigm of contextual cueing was used to investigate implicit learning of spatial context. The contextual cueing effect refers to the finding that in serial visual search tasks, the configuration of the target and the surrounding distractors can be learned implicitly, leading to faster target detection when configurations are repeated (Chun & Jiang, 1998). The first two experiments in this work demonstrated that contextual cueing did not occur in a pop-out task but at the transition from preattentive to attentive search. Further experiments investigated the robustness of contextual cueing. The contextual cueing effect was robust against large jitter of the configurations and did not depend on conspicuous arrangements in the configurations. In addition, an unvaried stimulus-response association did not enhance the learning effect. In the last section, the experiments aimed at the question whether the configurations could be explicitly learned. Results indicated that configurations could be memorized explicitly and it showed that in the course of time, explicit learning effects increased more than implicit learning effects. In the last experiment, this finding was replicated and extended for more repetitions of the configurations. To demonstrate that the contextual cueing effect is an implicit learning effect, results of a recognition test used by Chun and Jiang (1998) were replicated and an alternative recognition test was introduced which strengthened the former findings. Participants who were trained in the different tasks had remarkable learning effects of which they were not aware. Thus, results clearly demonstrated that contextual cueing is an implicit learning effect which is robust against a variety of manipulations.