The Impact of Mere Exposure Effect on Visual Stimuli

This experiment tests whether visual exposure positively alters preference for the test visual stimuli and secondly, the impact of repetitive exposure on the preference. The results support both the hypotheses - that conscious (supraliminal) and unconscious (subliminal) exposure increase liking towards the test visual stimuli and secondly, increasing the number of exposures positively correlates with affinity for the stimuli. The two hypotheses were tested using visible and suppressed geometric visual patterns. The data attained showed decreasing preference for the stimuli in the following order of exposure - conscious, unconscious and no prior exposure. The findings indicate that the impact is least in the non-exposed, less but still present in the subconscious exposure and most in the conscious, leading to the deduction that affinity for a stimulus occurs even without conscious cognition, the mere exposure effect, which is explained in greater detail below.

The first hypothesis is supported by Kunst-Wilson and Zajonc (1980), who tested the relationship between unconscious exposure and preference, and showed that positive reactions towards test stimuli were quicker for affect compared to cognitive, which led to the conclusion that liking is formed more affectively, and less cognitively.

The second hypothesis is consistent with studies by Newell and Shanks (2007), which conclude that greater exposure induces recognition, which in turn, increases familiarity and fosters a higher preference towards accustomed shapes. This concept is known as the mere-exposure effect (Zajonc, 1968). Results of this experiment can be attributed to the perceptual fluency model in which repeated and long exposure duration enhances overall fluency of processing visual stimuli and creates a preference effect (Newell & Shanks, 2007). Hence, both the hypotheses in this experiment are coherent with previous research, demonstrating that preference increases with greater exposure, regardless of the presence of conscious decision-making.

On the contrary, the correction process theory studied by Bornstein and D'Agostino (1992) found that the magnitude of mere exposure effect was greater in subliminal stimuli exposure rather than supraliminal when ratings were made with accuracy (above chance level). Bornstein and D'Agostino (1992) backed this theory by showing that subjects had engaged in the correction process, voluntarily discounting an otherwise greater affinity towards supraliminal stimuli. Our experiment is not the only experiment that did not concur with Bornstein and D'Agostino's findings, findings by Newell & Shank (2007) showed similar findings to ours.

The findings may be also attributed to several methodological limitations. Firstly, the specific usage of geometric visual patterns as stimuli may have induced a bias as preference for certain shapes appeals differently to individuals. Therefore, the adoption of other stimuli such as auditory (Heingartner & Hall, 1974), random shapes (Seamon, Marsh & Brody, 1984) and photographs of unfamiliar faces (Bornstein, Loene & Gallery, 1987) could possibly reduce this bias. Moreover, Wang and Chang (2010) suggest that preference is determined by the recollective state of individuals as opposed to the actual presence of stimuli. Hence, uncontrollable variable factors such as sleep deprivation (Havewoud et. al., 2010) and bored (Bornstein, Kale & Cornell, 1990), which impair cognitive processes and inhibit affective reaction, are limiting conditions for this experiment. Furthermore, our experiment required subjects to maintain a constant fixed level visual contact for optimal delivery of visual stimuli. Thus, a disruption of visual position at any time would compromise the result of the experiment. This study evaluated supraliminal and subliminal stimuli simultaneously; and due to this structure, preferences of subjects are susceptible to misattribution of perceptual fluency (Fang, Singh & Ahluwalia, 2007). Hence, residual preferences for supraliminal stimuli may have enhanced perceptual fluency when processing subliminal stimuli. Therefore to avoid this "contamination" of results of one by the other, two separate experiments should be performed one at a time, with either supraliminal or subliminal stimuli as the only variable, all other conditions