Measuring "intuition" in the SRT generation task.

We address some concerns related to the use of post-trial attribution judgments, originally developed for artificial grammar learning (AGL), during the version of the serial reaction time (SRT) task used by Fu, Dienes, and Fu (2010). In particular, intuition attributions, which are central to Fu et al.'s arguments, seem problematic: This attribution is likely to be made when stimuli contain several competing sources of information to which subjective feelings could be attributed. The interpretation of intuition attributions in Fu et al.'s SRT generation task is problematic because the procedure involved a 2-element sequence where items varied only in position. In our view, responses categorised as intuitions might have been a variety of guess response where neither judgement knowledge nor structural knowledge were conscious. The results would then be compatible with previous findings showing that people can control the use of unconscious structural knowledge even when judgement knowledge is unconscious.