Contextual Learning in Visual Search Tasks

Aim of the following six experiments is to test two aspects of the Contextual Cueing Effect: the relationship between attention and learning, and the attentional guidance account. Contextual Cueing can be defined as a strategy of the visual system to learn regularities from the environment to capitalize them. The main gain can be observed when considering reaction times in a visual search task: the target is detected faster when spatially related distractors are repeated.

Section 1 investigates thorough the components of the Contextual Cueing Effect and discusses some ambiguous results that emerge from several studies present in literature.

The two experiments discussed in Section 2 add evidence to the relationship occurring between dwell time and contextual learning, showing that the relationship between attentional resources and contextual learning does not increase monotonically. On the contrary, it is plausible assuming that the visual system needs a certain amount of time to learn the spatial contexts, and additional time does not increase the power of learning.

In Section 3 we will discuss the Contextual Cueing Effect that we observed in an efficient search task. We assumed that if the effect is achieved by a guidance mechanism, no contextual learning would emerge during an efficient search task. We show that the context effect can not always be observed during a pop-out search task; also, the different time course with respect to the classic contextual cueing effect and the weakness of the effect do not give us the possibility to make steady inferences about the attentional guidance account.


Section 4 presents an experiment that tests the attentional guidance hypothesis. We compared the slope of the RT x Set-size function with the patterns of eye movements in a gaze contingent contextual cueing experiment. This last experiment provides a strong evidence that contextual cueing can be achieved by an attentional guidance mechanism without the slope effect.