53 Chapter 3 people who adopt these reasons show a stronger impulse buying urge. We did not find a direct effect of message frames on impulse buying urge in Experiment 1. A possible explanation for the absence of this direct effect could be that consumers provide socially desirable answers, as impulse buying is generally linked to various negative traits, such as immaturity, and outcomes, such as financial problems (e.g., Rook & Fisher, 1995). Perhaps only people who explicitly mentioned moral justifications for buying the cookies when openly asked about their thoughts and feelings dared to indicate that they would impulsively buy the product. To determine whether the (socially responsible) answers given on the scale measuring impulse buying urge indeed differ from actual impulse buying behavior, we choose, in Experiment 2, to additionally measure impulse buying behavior (H2). Moreover, measuring impulse buying behavior in an actual shopping environment, as in Experiment 1, is challenging. In such cases, participants are often afterwards asked if the purchase was done impulsively (e.g., Mohan et al., 2013). This may lead to false answers, as it is likely that people defend their impulsively made choices by convincing themselves (and others) that the purchase was planned (see Moes et al., 2021). In Experiment 2, which will be a vignette-based experiment, we are able to measure actual impulse behavior without encountering this problem (see ‘Experiment 2’, ‘Measurements’). As we found, in Experiment 1, a two-step indirect effect of advertising frame on justification and of justification on impulse buying urge, we have reasons to believe that justification underlies the possible effect of message frames on impulse buying. Therefore, we test for mediation in Experiment 2 (H3). In Experiment 1, we compared self-benefit message frames with other-benefit message frames. To compare the effects of these message frames with a baseline, we added a neutral message frame to the design of Experiment 2. Experiment 2 Method. Design & participants. The second experiment (N = 524) was a vignette-based experiment in which participants were explicitly asked to imagine performing groceries in a physical supermarket (cf. Atzmüller et al., 2010; Verhagen et al., 2019). The aim of Experiment 2 was to 1) examine whether there is an effect of message frame on impulse buying urges (H1) and on impulse buying behavior (H2), and 2) test whether the most commonly reported justifications, derived from the open-ended questions in Experiment 1 (cf. Mukhopadhyay & Johar, 2009; Taylor, Webb & Sheeran, 2014), functions as a mediator in the effect of message frames on impulse buying (H3). Experiment 2 used a single-factor between-subjects design with “message frame” (self-benefit versus other-benefit versus neutral message frame) as the IV, “justification” as mediating variable, and “impulse-buying urge” and “impulse buying behavior” as the DVs. We hired a data collection bureau to recruit participants throughout the Netherlands. From the 1039 people who started this online experiment, 524 passed the two attention checks and fully completed the questionnaire. The sample was representative of the Dutch population regarding sex (50.4% identified as female, 49.4% as male, and 0.2% neither identified as male nor female), age (M = 47.61, SD = 16.25), and education (29%
RkJQdWJsaXNoZXIy MjY0ODMw