Measuring unconditional stimulus expectancy during evaluative conditioning strengthens explicit conditional stimulus valence

ABSTRACT

During evaluative conditioning, a neutral conditional stimulus (CS) becomes pleasant or unpleasant after pairings with a positive/negative unconditional stimulus (US). Measures of US expectancy are commonly assessed during conditioning but it is unclear whether this affects evaluative learning. In Experiment 1, we examined whether the concurrent assessment of US expectancy alongside measures of CS valence would influence the acquisition, extinction, and reinstatement of explicit CS valence. Participants rated both valence and expectancy during conditioning (valence/expectancy group) or only CS valence (valence only group). Evaluative conditioning was acquired in both groups during acquisition, but its magnitude was enhanced in the valence/expectancy group. Measuring US expectancy did not influence the extinction or reinstatement of conditional valence. In Experiment 2, we confirmed the enhancement of evaluative conditioning due to concurrent measurement of US expectancy in an explicit measure, but did not find corresponding evidence in an implicit measure of conditional valence. In Experiment 3, we replicated the results using a different US expectancy scale and demonstrated that measuring CS valence multiple times throughout conditioning also strengthens conditional valence. Overall, the results suggest that the measurement of US expectancy and CS valence throughout conditioning draws attention to the contingencies and strengthens explicit evaluative learning.

KEYWORDS:

• Evaluative conditioning
• valence
• US expectancy measures
• reinstatement
• CS valence measures

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data from all experiments can be found at https://osf.io/xzsca/.

Notes

1 The multivariate solution is used as it has been shown to be more robust against assumption violations (see Vasey & Thayer, 1987).

2 This marginal main effect of group reveals that overall evaluations (collapsed across CSp and CSu) are more unpleasant in valence/expectancy group.

Additional information

Funding

This work was supported by grants number DP180100869 and SR120300015 from the Australian Research Council.