Abstract: This paper studies the effect of self-construal on consumers’ purchase intention of products that vary in visual novelty, its underlying mechanism and boundary conditions. Specifically, we propose that products with low (high) visual novelty should increase the purchase intention among interdependent (independent) consumers through decreased social risk perception (increased consumer needs for uniqueness). In addition, the interaction between visual novelty and self-construal should be reduced when product type (hedonic and utilitarian) is made salient. This is because all consumers should prefer utilitarian products with low visual novelty but prefer hedonic products with high visual novelty, regardless of self-construal. Three experiments provide empirical support for these predictions.
Study 1 uses 2 (visual novelty: low/high) × 2 (self-construal: independent/interdependent) between-subjects design. A night-light with an ambiguous product type is selected as the stimulus in the form of a color print advertisement. We recruit 112 participants, design different models, distribute product pictures with different appearances, and use an existing scale to measure participants' self-construal. We confirm the interaction between visual novelty and self-construal on purchase intention.
Study 2 uses a similar between-subjects design, and a wrist-watch with an ambiguous product type as the stimulus. We recruit 140 participants and manipulate the product’s visual novelty and self-construal. Consistent with our prediction, we find an interaction between self-construal and visual novelty on consumers’ purchase intention. In addition, we provide supporting evidence for the proposed mechanism for the interaction effect that is due to consumer needs for uniqueness and perceptions of social risks.
Study 3 uses a 2 (visual novelty: low/high) × 2 (self-construal: independent/interdependent) x 2 (product type: hedonic/utilitarian) between-subjects design, to test the moderating effect of product type. Product visual novelty and self-construal are manipulated in similar manners as in Study 2. Desk-lamp is used as the stimulus. In order to minimize confounds, we manipulate the product type and verify our manipulation in a pretest. In another pretest we also verify the effectiveness of the visual novelty manipulation and rule out the potential confound of functional novelty. We recruit 302 participants for this study. The results provide support for the moderating effect of product type. Specifically, we find that, regardless of self-construal, all consumers prefer a utilitarian product with lower appearance novelty, but prefer a hedonic product with higher appearance novelty. In addition, these effects are mediated by consumer needs for uniqueness and perceptions of social risks.
In summary, the results of three studies provide convergent evidence for an interaction between self-construal and product visual novelty on consumers' purchase intention, the moderating effect of product type, and the underlying mechanism due to consumer needs for uniqueness and social risk perceptions.

Key words: visual novelty, self-construal, need for uniqueness, social risk perception, purchase intention