The effects of stimulus type in aesthetic preference research : a meta-analysis

Using meta-analysis, this study investigated the effects of different stimulus types (analytic and synthetic) on outcomes of aesthetic preference studies. Analytic stimuli are actual art objects. Synthetic stimuli are nonart visual displays. These have been assumed to be equivalent in research. If this assumption were correct, study outcomes, expressed as effect sizes, would comprise a statistically homogeneous group. Instead, this study found that outcomes grouped according to stimulus type were heterogeneous. Further analysis revealed no variables which accounted for this heterogeneity, precluding inferences regarding the precise effects of stimulus type difference. Nevertheless, descriptive statistics revealed that in the studies included, the two stimulus types elicited different mean aesthetic preference responses. Experimental studies using analytic stimuli had a mean ES of 0.774. Those using synthetic stimuli had a mean ES of 0.520. Correlational studies using analytic stimuli had a mean ES of 0.340. Those using synthetic stimuli had a mean ES of 0.420. The extent and direction of the difference interacted with other variables, including publication type, study design, subject-centered variables, and stimulus object-centered variables. It was concluded that the assumption that the different stimulus types are equivalent is suspect. It appeared that stimulus type difference was associated with differences in the outcomes of studies of aesthetic preference. It was recommended that clarification of aesthetic experience be pursued by expanding inquiry to include cognitive and qualitative perspectives, particularly since instruction in aesthetics has been given credence by the Discipline Based Art Education movement. Also, future studies of aesthetic preference should be designed so as to control for possible differential effects of stimulus type on outcome. This recommendation should also apply to interpreting previous research.