Abstract
We have recently shown that typical results from the unconscious priming paradigm suffer from serious methodological flaws that call into question the conclusions of many unconscious priming studies. Here, we describe a similar problem in the contextual cueing literature. In the priming paradigm, participants perform close to chance when directly attempting to identify or discriminate prime stimuli (direct task). Nevertheless, the same stimuli indirectly affect responses (e.g., priming effects on reaction times). From this pattern of results, researchers typically infer better sensitivity for the stimuli in the indirect than direct task, which is further interpreted as processing outside of conscious awareness. However, we have shown that this inference is flawed (Meyen et al., 2020). Similar to the priming paradigm, the contextual cueing paradigm follows the same logic: Participants show close to chance sensitivity in a direct task. In an indirect task, clear cueing effects on reaction times are found. This pattern of results is again used to infer better sensitivity in the indirect than direct task, which is then the basis for further-reaching inferences about unconscious processing. We generalized our methods to the typical contextual cueing paradigm (which often utilizes more response alternatives than the priming paradigm) and estimated the sensitivities of responses in the direct and indirect tasks based on the typically reported results. To do so, we present a model that relates participants’ sensitivity in the direct task to the size of cueing effects in the indirect task. This allows us to compare sensitivities between the two tasks and test for a difference, which is required to make further inferences on unconscious processing. We reanalyzed multiple studies showing that some fail to provide evidence for better sensitivity in the indirect task, thereby calling into question the corresponding inferences about conscious/unconscious processing.