Does Peer Buyer Attractiveness in Customer-Generated Images Impact E-Commerce Apparel Purchase Intentions?
Description
Customer-generated images (CGIs) on e-commerce platforms drive consumer decisions, but their precise impact on purchase intention remains under-explored, especially considering diverse levels of peer-buyer attractiveness (PBA) in apparel CGIs. To investigate this, we conducted five experiments involving female apparel shoppers (N1a = 230, N2a = 240, N2b =252, N3 = 284) and undergraduates (N1b = 95), employing experimental-causal-chain and measurement-of-mediation designs. Study 1 assessed PBA's effect on purchase intention, Study 2 explored self-referencing as a mediator, and Study 3 comprehensively evaluated a chain-mediation model. Hypotheses: Hypothesis 1: CGIs negatively influence purchase intention (a) and CGIs with low PBA lead to lower purchase intention than those with high PBA (b). Hypothesis 2: CGIs positively influence self-referencing (a) and CGIs with low PBA lead to lower self-referencing levels than those with high PBA (b). Hypothesis H3: Self-referencing mediates the effect of PBA on purchase intention. Findings revealed that medium and low PBA in CGIs reduced purchase intentions, mediated by self-referencing. CGIs offer emotional benefits by alleviating body image distress; low PBA proves most effective, while medium PBA is less so. They also provide cognitive advantages by reducing uncertainty consistently across all PBA levels. Notably, chain-mediating effects of self-referencing and body image distress, also self-referencing and uncertainty, were identified between PBA and purchase intention. Hypothesis 4: CGIs alleviate body image distress (a); compared with those with high PBA, CGIs with medium PBA lead to higher body image distress (b), while CGIs with low PBA lead to lower body image distress (c). Hypothesis 5: Body image distress mediates the relationship between PBA and purchase intention. Hypothesis 6: Self-referencing and body image distress plays a chain-mediating role between PBA and purchase intention. Hypothesis 7: CGIs reduce perceived uncertainty (a); CGIs with low PBA are associated with higher uncertainty than those with medium- or high PBA (b). Hypothesis 8: Uncertainty plays a mediating role in PBA's impact on purchase intention (PI). Hypothesis 9: Self-referencing and uncertainty play a chain-mediating role between PBA and purchase intention. Findings revealed that medium and low PBA in CGIs reduced purchase intentions, mediated by self-referencing. CGIs offer emotional benefits by alleviating body image distress; low PBA proves most effective, while medium PBA is less so. They also provide cognitive advantages by reducing uncertainty consistently across all PBA levels. Notably, chain-mediating effects of self-referencing and body image distress, also self-referencing and uncertainty, were identified between PBA and purchase intention.
Files
Steps to reproduce
For experimental manipulation, we selected a popular pair of training leggings, a basic T-shirt, and a pair of wide-legged jeans, all of which ranked in the top 5 in their respective categories on Taobao during the experiment week. Real customer review photos were chosen to comprehensively display garments worn by previous buyers. Eight undergraduates participated in a focus group to select images for further editing based on specific criteria: (1) high clarity suitable for image modification; (2) a nonprofessional pose by an actual customer, not a professional or semiprofessional model; and (3) a gym background for the training leggings. The gym was selected because it contains fewer socioeconomic status and aesthetic taste cues compared to other popular backgrounds such as homes, the outdoors, or tourist attractions. (The T-shirt and jeans followed the same criteria, but with fitting rooms as the background). Three CGIs were chosen as the prototype. To manipulate high, medium, and low levels of physical attractiveness, the prototype CGI was modified using Adobe Photoshop software. The modifications adjusted body proportions, shapes, and lines, following Feng's (2019) experimental study. Buyers' faces were obscured to eliminate potential facial effects (Guan et al., 2023; Li, Zhang, et al., 2022). To simulate real-world reviews, the modified CGIs also displayed the buyer's ID (labeled as "anonymous"), a concise positive text review on the product's performance, and the count of comments and likes for the review (both set to 0). In summary, for each product, three sets of experimental materials were produced. Each set comprised three versions of the same CGI, manipulated to vary only in physical attractiveness while ensuring consistency in other factors like background, model pose, image quality, and other review-related information. Pretests were conducted, with refinements based on their results to ensure optimal experimental stimuli. The final pretest validated the CGI's use for PBA manipulation because ANOVA revealed significant differences in buyer attractiveness ratings across the CGI versions: low-PBA < medium-PBA < high-PBA (ps < .05). No significant variations were observed for trustworthiness or model attractiveness. Seller-buyer discrepancy ratings also significantly differed among groups (p < .05), with participants primarily attributing the discrepancy to characteristics rather than the product.