Is task-irrelevant learning really task-irrelevant?

Abstract

In the present study we address the question of whether the learning of task-irrelevant stimuli found in the paradigm of task-irrelevant learning (TIPL) [1]-[9] is truly task irrelevant. To test the hypothesis that associations that are beneficial to task-performance may develop between the task-relevant and task-irrelevant stimuli, or the task-responses and the task-irrelevant stimuli, we designed a new procedure in which correlations between the presentation of task-irrelevant motion stimuli and the identity of task-targets or task-responses were manipulated. We found no evidence for associations developing between the learned (task-irrelevant) motion stimuli and the targets or responses to the letter identification task used during training. Furthermore, the conditions that had the greatest correlations between stimulus and response showed the least amount of TIPL. On the other hand, TIPL was found in conditions of greatest response uncertainty and with the greatest processing requirements for the task-relevant stimuli. This is in line with our previously published model that suggests that task-irrelevant stimuli benefit from the spill-over of learning signals that are released due to processing of task-relevant stimuli.

Figure 1. Design of Experiment.

Tests, subjects conducted tests before and after training in which they reported the direction of motion coherence by selecting an arrow with a computer mouse. Training, a given trial was from one of three conditions; in the exact-target condition the targets were always the same for a given subject (here <3 2>), in the target-present trials the targets could be any number combination other than that presented in the exact-target condition, and in the target-absent condition no target was presented were to respond . Red arrows indicate direction of coherence; a different direction of motion coherences was paired with each trial type.

Figure 2. Performance change on the direction discrimination task, between pre- and post tests.

Unexposed directions data represent the average across the 3 directions that were not exposed during the training phase. Error bars represent within-subject standard error ; see methods for details.

Figure 3. Results from the RSVP task performed during the training sessions.

Plots represent data from target-present (left), exact-target (middle) and target-absent (right) trials, respectively. The break in the lines represents that the data from the first 7 days are from 3 subjects and that of last 10 days are from all 7 subjects. Each line represents trials in which a different direction was presented; exact-target direction (blue), target-present direction (green), target-absent direction (red). Error bars represent within-subject standard error [20]).