"Should I stop or should I go? The role of associations and expectancies": Correction to Best et al. (2016).

Abstract

Reports an error in "Should I stop or should I go? The role of associations and expectancies" by Maisy Best, Natalia S. Lawrence, Gordon D. Logan, Ian P. L. McLaren and Frederick Verbruggen (Journal of Experimental Psychology: Human Perception and Performance, 2016[Jan], Vol 42[1], 115-137). In the article, there is an error in Table 3 of the Results and third paragraph of the Results section labeled Test phase. In Experiment 4, the study performed an exploratory post-hoc test of the go reaction times in the training phase, contrasting stop-associated and go-associated items. Control items were excluded. Instead of reporting the results of the full analysis (with all three items types included), the authors incorrectly reported the results of this post-hoc analysis in Table 3 and in the main text. The correct analysis is presented below. Note that all other analyses reported in the tables and main text are correct. The R code shared via Open Research Exeter data repository (http://hdl.handle .net/10871/17735) is also correct. The interaction between image type and block is no longer significant when control items are included (p .094; p .037 for the post-hoc test). (The following abstract of the original article appeared in record 2015-40003-001.) Following exposure to consistent stimulus-stop mappings, response inhibition can become automatized with practice. What is learned is less clear, even though this has important theoretical and practical implications. A recent analysis indicates that stimuli can become associated with a stop signal or with a stop goal. Furthermore, expectancy may play an important role. Previous studies that have used stop or no-go signals to manipulate stimulus-stop learning cannot distinguish between stimulus-signal and stimulus-goal associations, and expectancy has not been measured properly. In the present study, participants performed a task that combined features of the go/no-go task and the stop-signal task in which the stop-signal rule changed at the beginning of each block. The go and stop signals were superimposed over 40 task-irrelevant images. Our results show that participants can learn direct associations between images and the stop goal without mediation via the stop signal. Exposure to the image-stop associations influenced task performance during training, and expectancies measured following task completion or measured within the task. But, despite this, we found an effect of stimulus-stop learning on test performance only when the task increased the task-relevance of the images. This could indicate that the influence of stimulus-stop learning on go performance is strongly influenced by attention to both task-relevant and task-irrelevant stimulus features. More generally, our findings suggest a strong interplay between automatic and controlled processes. (PsycINFO Database Record