Extinction of likes and dislikes: effects of feature-specific attention allocation

ABSTRACT

The evaluative conditioning (EC) effect refers to the change in the liking of a neutral stimulus (conditioned stimulus, CS) due to its pairing with another stimulus (unconditioned stimulus, US). We examined whether the extinction rate of the EC effect is moderated by feature-specific attention allocation. In two experiments, CSs were abstract Gabor patches varying along two orthogonal, perceptual dimensions (i.e. spatial frequency and orientation). During the acquisition phase, one of these dimensions was predictive of the valence of the USs. During the extinction phase, CSs were presented alone and participants were asked to categorise the CSs either according to their valence, the perceptual dimension that was task-relevant during the acquisition phase, or a perceptual dimension that was task-irrelevant during the acquisition phase. As predicted, explicit valence measures revealed a linear increase in the extinction rate of the EC effect as participants were encouraged to assign attention to non-evaluative stimulus information during the extinction phase. In Experiment 1, Affect Misattribution Paradigm (AMP) data mimicked this pattern of results, although the effect just missed conventional levels of significance. In Experiment 2, the AMP data revealed an increase of the EC effect if attention was focused on evaluative stimulus information. Potential mechanisms to explain these findings are discussed.

KEYWORDS:

• Evaluative conditioning
• feature-specific attention allocation
• extinction
• selective attention

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. As pointed out by an anonymous reviewer, this procedure is similar to the procedures used in research on the so-called intradimensional–extradimensional shift effect (see George & Pearce, 1999). In line with the FSAA framework, it has been argued that the learning rate in an operant condition paradigm is higher for stimulus dimensions that are relevant (as compared to irrelevant) for training discriminations during a preceding training phase (see Sutherland & Mackintosh, 1971). Nevertheless, there is some experimental evidence that argues against the existence of this effect (Trobalon et al., 2003). For an extensive and critical discussion, see Kattner and Green (2016).

2. There is some debate concerning the extent to which AMP effects are driven by participants who rate the primes intentionally (see Bar-Anan & Nosek, 2012). We examined this possibility using the procedures described by Payne et al. (2013). In both experiments, AMP scores did not differ between participants who claimed that they had rated the primes intentionally and participants who did not claim that they had rated the primes intentionally, all t’s < 1.

3. When we included the 10 participants who did not show an EC effect following acquisition, the valence ratings still revealed a significant effect of condition, F(1, 88) = 4.07, p < .05, η² = .04. Significant effects were absent in the AMP, F(1, 88) = 1.32, p = .25, η² = .01, and the expectancy ratings, F < 1.

4. To examine whether the EC effect was dependent upon (explicit) US expectancy, valence ratings were subjected to a linear mixed effects model. Fixed effects were CS type (positive vs. negative), US expectancy (in line with actual pairings vs. not in line with actual pairings), and the interaction between these factors. Participants and stimuli were defined as crossed random effects. The mixed-model F-tests were computed using Kenward–Roger’s adjusted degrees of freedom (Kenward & Roger, 1997). The analysis revealed a strong in interaction between CS and US expectancy both in Experiment 1, F(1, 1486.07) = 252.15, p < 0.001, and in Experiment 2, F(1,1526.02) = 102.43, p < 0.001. In Experiment 1, however, a significant EC effect emerged irrespective of whether the US expectancy ratings were in line with the actual CS–US pairings, F(1, 1197.26) = 785.32, p < 0.001 or were not in line with the actual CS–US pairings, F(1, 241.93) = 24.18, p < .001. In Experiment 2, a reliable EC effect was found only if the US expectancy ratings were in line with the actual CS–US pairings, F(1, 1297.57) = 1142.80, p < 0.001.

Additional information

Funding

This work was supported by the Methusalem Grant BOF16/MET_V/002 of Ghent University awarded to Jan De Houwer.