Be careful what you say! – Evaluative change based on instructional learning generalizes to other similar stimuli and to the wider category

ABSTRACT

In evaluative conditioning, a conditional stimulus (CS; e.g. a neutral picture) acquires positive/negative valence if it is paired with a pleasant/unpleasant unconditional stimulus (US; e.g. a positive/negative picture). This valence generalises to other stimuli similar to the CS and to the wider CS category. Being informed that the CS will be paired with the US induces a similar change in valence (instructional learning), but it is not clear whether instructional learning would also generalise. In Experiment 1, participants were informed that one shape would be paired with pleasant and another with unpleasant images. These instructions instilled conditional valence to the CSs which generalised to different shapes from the same category (generalisation stimuli). In Experiment 2, we replicated this finding in an implicit measure using stimuli varying in perceptual features. Participants were informed that three CSs from one category (e.g. vegetables) would be paired with pleasant images and three CSs from a second category (e.g. office supplies) would be paired with unpleasant images. This instruction instilled conditional valence to the CSs which generalised to novel exemplars from the same categories. This suggests that conditional valence instilled via instructions generalises to other stimuli – a finding with implications for prejudice and racism.

KEYWORDS:

• Evaluative conditioning
• conceptual generalisation
• instructed acquisition
• instructional learning
• valence

Acknowledgements

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

Disclosure statement

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

Data availability statement

Data for all experiments can be found at https://osf.io/rwgbc/.

Notes

1 This interaction attains significance in participants who passed the instruction manipulation check, F(1, 158) = 4.03, p = .046, ηp² = .025.

Additional information

Funding

This work was supported by [grants numbers DP180100869 and SR120300015] from the Australian Research Council.