Subliminal influence on preferences? A test of evaluative conditioning for brief visual conditioned stimuli using auditory unconditioned stimuli

Abstract

In the field of evaluative conditioning (EC), two opposing theories—propositional single-process theory versus dual-process theory—are currently being discussed in the literature. The present set of experiments test a crucial prediction to adjudicate between these two theories: Dual-process theory postulates that evaluative conditioning can occur without awareness of the contingency between conditioned stimulus (CS) and unconditioned stimulus (US); in contrast, single-process propositional theory postulates that EC requires CS-US contingency awareness. In a set of three studies, we experimentally manipulate contingency awareness by presenting the CSs very briefly, thereby rendering it unlikely to be processed consciously. We address potential issues with previous studies on EC with subliminal or near-threshold CSs that limited their interpretation. Across two experiments, we consistently found an EC effect for CSs presented for 1000 ms and consistently failed to find an EC effect for briefly presented CSs. In a third pre-registered experiment, we again found evidence for an EC effect with CSs presented for 1000 ms, and we found some indication for an EC effect for CSs presented for 20 ms.

No evaluative conditioning effects with briefly presented stimuli

“…Taken together, Heycke et al (2017) found an EC effect for briefly presented -but slightly supraliminal -CSs in a cross-modal presentation paradigm. While the cross-modal approach appears promising, it remains to be seen whether the EC effect found by Heycke et al (2017) can be replicated, and whether it relies on above-chance CS identification or extends to CSs detected at chance level.…”

“…As the current study aims at replicating the findings by Heycke et al (2017), we used the same experimental script, with some minor adjustments: (1) In order to ensure that the effect found by Heycke et al (2017) can be generalized to different stimuli, we used a new set of CSs that were randomly assigned to a brief and a longer CS presentation time condition. (2) We additionally added an extensive visibility check, which was conducted after the measurement of the dependent variable.…”

“…One could speculate, however, that this online visibility check might interfere with the learning procedure in two ways: (I) participants might be in a deliberative and analytic mindset during the task as they are trying to identify the briefly presented stimuli, and therefore associations between the briefly presented CS and the US might not form if they depend on holistic processes, (II) assuming that a new evaluative response toward the CSs is acquired during the learning phase, showing multiple CSs that were paired with USs of different valence after every trial during the learning phase could be considered an additional trial of the learning phase, and could therefore interfere with the intended CS-US pairing schedule. A recent experiment by Heycke, Aust, and Stahl (2017) attempted to tackle this potential interference by omitting the online visibility check, while also addressing another potential limitation: the presentation schedule of the CS-US pairs.…”

Putting Operating Principles (What) Before Conditions (How and When) to Improve Theorizing: S - S versus S - R learning in Evaluative Conditioning

“…For example, a researcher might find that in one type of procedure, or among a certain class of participants, or with a certain kind of stimuli, EC can be established by means of subliminal presentations, while it cannot with other procedures, participants or stimuli (Heycke, Aust, & Stahl, 2017). As long as we have no further information on EC's operating principles in each setting, these findings, while interesting, would remain puzzling pieces of conflicting information.…”

The past, present, and future of Registered Reports

“…RRs in many journals can include preliminary studies (e.g. Heycke et al 77 ), prespecified sequences of studies (e.g. Ait Ouares et al 78 ) or incremental registrations after Stage 2 acceptance.…”