The relationship between affect and recognition memory for reproductions of paintings
Seven experiments were performed with complex visual stimuli (color reproductions of paintings). Subjects were tested with Yes/No recognition memory. In Experiment 1, a levels of processing manipulation yielded better recognition performance for single color and monochrome pictures encoded with a liking question than with a color-of-stimulus question. In Experiment 2, subjects studied a new set of single color pictures restricted to Landscapes, Portraits and Still Lifes. Recognition conditionalized on liking (using a five-point scale) revealed a modest "Pollyanna" effect: liked items were best recognized.
In order to manipulate affect, experimenter controlled paired stimuli were created from Experiment 2 pictures by factorially combining Categorical, conceptual relatedness (Related: Portrait/Portrait, and Unrelated: Landscape/Still Life) and Aesthetic, stylistic relatedness (Related and Unrelated: established by ratings). Subject ratings established that among the four Pair Types, Categorically and Aesthetically Related pairs were highest rated on all types of relatedness, followed by Categorically Related, Aesthetically Unrelated pairs and Categorically Unrelated, Aesthetically Related pairs; and last, Categorically and Aesthetically Unrelated pairs. In Experiments 3-6, the paired stimuli were incidentally or intentionally encoded, sometimes with aesthetic relatedness encoding tasks.
Results across experiments revealed that Categorically Related pairs yield better mean recognition performance, but Aesthetically Related pairs are preferred, even when lacking categorical relatedness. Categorically and Aesthetically Related pairs were consistently best recognized and most liked. Portraits as a group were sometimes better recognized, but this effect was not the source of significant main effects. In Experiment 3, a Pollyanna effect was found for pairs, and in Experiment 6, a context change manipulation produced diminished performance for changed context items, in particular for Aesthetically Related pairs, suggesting that aesthetic affect may function somewhat like mood.
For single picture stimuli, the impact of affect on memory seems direct and preference-based, although it is unclear if subjects experience aesthetic affect. Paired picture stimuli make complex affective and information processing demands, and better recognition for them may be predominantly a function of facilitation of encoding when pairwise information is semantically and visually redundant.
0273: Art education