Study 1 - A total of 410 women were included in the data analysis. The average height of the participants was 162 cm, ranging from 105 cm to 183 cm. The average weight was 70.82 kg, ranging from 40 kg to 122 kg. The participants were invited to participate in a survey about a women's fashion brand that would be launched soon in the Brazilian market and that the survey would be used to outline the profile of potential consumers of this new brand. First, the participants answered questions about their fashion profile, shopping habits, and clothing size. Next, they randomly saw one of the two ads featuring standard or plus-size model (M-VGap manipulation), and answered questions related to BInt and AtoB. Finally, the participants were asked about their physical features as height, weight, and BSat/BDiss. Study 2 - A total of 729 participants were included in the data analysis. The average height of the participants was 162 cm, ranging from 104 cm to 192 cm. The average weight was 72.54 kg, ranging from 41 kg to 129 kg. Different from Study 1, the models portrayed in the ads for the manipulation of M-VGap were black women. In the field work, the procedure was identical to that of Study 1.
Steps to reproduce
M-VGap was operationalized by the difference between the Body Mass Index (BMI) of the model and the BMI of the viewer. The first one was determined based on the results of a pre-test with 30 potential respondents and on parameters previously stated in literature. Participants were asked to evaluate the BMI of the models selected for the Studies by using the images scale from Pulvers (2013). Then, the BMI for each model was classified by the relative body dimensions from Kakeshita & Almeida (2006) . The BMI of the participants was calculated using their self-reported weight (kg) and height (cm). The M-VGap was calculated by the difference between the BMI of the model and the BMI of the viewer. The lower the gap is (negative), the thinner the model in the ad is than the viewer; the higher the gap is (positive), the heavier the model in the ad is than the viewer. BSat/BDiss was operationalized by two questions on the body size Pictorial Body Image Scale (1 = extremely thin, 9 = extremely heavy) (Pulvers, 2013). The questions were: 'Among the silhouettes below, at this moment I see myself as the silhouette number...'; 'Among the silhouettes below, I would like to be like the silhouette number...'. Then we compared the two silhouettes informed by the participants to measure their BSat/Diss. We consider that if the self-silhouette matches the wished silhouette the respondent is satisfied with his/her body. BSat decreases as the self and wished silhouettes move away. This means that when the value is zero, BSat exists and BDiss does not exists. All other values for BSat represent some level of BDiss. The dependent variables Attitude towards the brand and Purchase Intention were measured by previous reported scales. To manipulate the M-VGap we prepared two ads for a new women's fashion brand. To select the models of the ads, a total of 49 females participated in the pre-test. The photographs of 20 models from a real online apparel store were analyzed. We selected 10 pairs of photographs, one portraying a standard model and one portraying a plus-size model, both wearing the same clothes. For the 10 pairs of pics, the participants answered questions about how they perceived each model: as a plus-size model or as a standard model. They also answered questions about other characteristics of the model: beauty, intelligence, attractiveness, good look, happiness, satisfaction with one's own body, healthy and whether they looked like professional models or not. The two pairs of models chosen for Studies 1 and 2 were recognized by participants as a standard model and a plus-size model and kept the same level of evaluation in the other measured characteristics. Then, using a background image as a template, we created two full-color ads for each experiment . The only difference across the two ads was the model. Other creative elements, such as background, copy, and imagery, were identical across the two stimuli.