Evaluative conditioning with fear- and disgust-evoking stimuli: no evidence that they increase learning without explicit memory

ABSTRACT

Evaluative conditioning (EC) is a change in the liking of a stimulus (conditioned stimulus or CS) due to its previous pairings with another stimulus (unconditioned stimulus or US). In three experiments, we investigated if conditioning CSs with fear or disgust evoking USs increase EC effects that do not require explicit memory. Experiment 1 used images to manipulate the type of US between participants, Experiment 2 used auditory stimuli to manipulate the type of US within participants, and Experiment 3 used both images and auditory stimuli to manipulate the type of US within participants. All experiments failed to provide evidence that fear/disgust-evoking USs lead to larger EC effects without explicit memory than non-fear/non-disgust-evoking USs. Most results are in line with the assumption that the found EC effects are based on explicit memory. The results contribute to the larger goal in EC research of determining which conditions can lead to EC effects in the absence of explicit memory.

KEYWORDS:

• Evaluative conditioning
• fear-evoking
• disgust-evoking

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 We preregistered in Experiments 1 and 2 a Multinomial Processing Tree (MPT; see Hütter et al., 2012) to disentangle EC effects based on memory versus attitudes without memory. However, it became apparent that parameter estimates were unreliable due to the design and outcomes of the conditions. On one hand, the design did not allow us to analyse conditions of valence and negative US-type separately. On the other hand, as pointed out to us during the review process, estimating parameters for fear/disgust-evoking and non-fear/non-disgust-evoking conditions jointly for positive and negative valences led to unreliable parameter estimates, as could be inferred from different frequencies in these conditions. Unfortunately, the problem is difficult to overcome by changing the design in a way that would allow separate estimations of valence and fear/disgust-evoking: Positive stimuli simply cannot be manipulated on the fear/disgust-evoking dimension in a meaningful way. We thus refrained from reporting our MPT results, and we instead fit linear mixed models to test our hypothesis. We planned Experiment 3 with a different analytical procedure in mind.

2 Positive sounds in the fear/disgust-evoking condition do not evoke fear/disgust themselves, but are referred to as fear/disgust-evoking based on the design.

Additional information

Funding

This research was supported by grant GA 1520/2-1 from the Deutsche Forschungsgemeinschaft awarded to Anne Gast.