Fixation and attention control in lateralised target detection and free recall with words

Abstract

The purpose of the present study was to test the hypothesis that lateralised target detection in the visual modality would produce results similar in magnitude, reliability, and validity to those obtained in the auditory modality with an analogue task. Thus, it was expected that it would produce laterality effects that are larger, more reliable, and more valid than those obtained in a free recall task. The claim that target detection provides its own attention control also led to the hypothesis that the magnitude of laterality effects should be affected by fixation control in free recall but not target detection. A total of 349 right-handed participants completed a word recognition task with either free recall or target detection with or without fixation control. Only the finding that free recall was generally more reliable than target detection went contrary to the hypotheses. This finding is interpreted as reflecting a consistent attentional bias that stems from task requirements. In general, the results suggest that target detection without fixation control has much potential as a measure of perceptual asymmetries in the visual modality.

Acknowledgements

This study was funded by the Natural Sciences and Engineering Research Council of Canada to D. Voyer. The authors thank Michael Celozzi, Michael Dillard, Merrilee Egers, Diana MacLean, Jeff Phillips, Adam Schwartz, Mariana Soraggi, and Stephanie Williams for their assistance with data collection and scoring.

Notes

¹It is plausible to argue that the ratio to chance performance as calculated by Voyer (2003) might be problematic. Specifically, if, in some extreme hypothetical example, chance performance was 90%, then the maximum ratio that could be obtained is 1.11, but if chance performance was 10%, then the maximum would be 10.00. Thus, the maximum value of such a ratio score is proportional to the underlying chance level. In order to provide some control of this problem, a value expressing performance as a ratio of what is achievable was calculated for each participant in the form (X-chance)/chance. Data were analysed with this new ratio as the dependent variable and the results produced the same significant findings as the ratio based on Voyer's (2003) approach. This was possibly due to the fact that chance performance was relatively similar in free recall (33.3%) and target detection (50.0%). Accordingly, the results presented in what follows preserved Voyer's ratio to allow comparison with his work.

²A participant with perfect knowledge who perceived that “yes” responses were required 2/3 of the time, could actually achieve 56% accuracy by responding randomly at that rate (i.e., 2/3 of the 2/3 target present trials, plus 1/3 of the 1/3 target absent trials). However, in practice such knowledge could only be acquired in the course of task performance, which would defer if not prevent its acquisition. Alternatively, if “yes” responses were emitted 100% of the time, 67% accuracy could be obtained (i.e., all of the 2/3 target present trials). However, in this case accuracy would be 0% on the target absence trials, and that was never observed: In fact accuracy averaged 55.7% on those trials. The assumption of a 50% chance rate therefore appears reasonable to us and by any means of reckoning should not be inaccurate to any great degree.