56 The Urge to Splurge buy, which are communicated through an other-benefit message frame (compared to a self-benefit message frame) and use them to morally justify impulsively buying the cookies. We show that people seem to need altruistic reasons to buy an indulgent product on impulse and that sustainability-driven companies can provide these reasons to consumers, which is of high value for practitioners. We performed a sensitivity analysis, using G*Power (Faul et al., 2007) to assess the minimal effect size Experiment 2 could reliably detect, given the used sample size (N= 524). This analysis showed that the smallest effect size to be detected was f2 = .02, which was substantially smaller than the found effect size (f2 = .17). Therefore, this experiment appeared to be well powered to detect effect sizes in the small to medium range, thus limiting type 1 error concerns. Nevertheless, we decided to replicate Experiment 2 for robustness and reliability. Replication is essential to ensure certainty in the found results (McEwan et al., 2018). Experiment 3 Method. Design, participants, stimulus material, & procedure. The third experiment (N = 527) was also vignette based. The aim of Experiment 3 was to replicate the results observed in Experiment 2 and thus again test Hypotheses 1– 3, as well as to test the effect of impulse buying urges on impulse buying behavior (H4) to present a complete conceptual model. Experiment 3 is an exact replication of Experiment 2, with the exception of the few minor changes that we describe below. Experiment 3, similar to Experiment 2, used a single-factor between-subjects design with “message frame” as the IV. However, as the neutral message frame did not significantly differ from the self-benefit and other-benefit message frames in Experiment 2, we only tested the two latter conditions in Experiment 3 (self-benefit versus other-benefit message frame). This way we created a plain replication of the observed effects. From the 1096 people who started the online experiment, 527 passed the two attention checks and fully completed the questionnaire (nother = 237; nself = 290). The sample was again representative of the Dutch population vis-à-vis sex (51.4% identified as female, 48.4% as male, 0.2% did not identify as male or female), age (M = 48.13, SD = 16.68), and education (31.7% was theoretically trained, 35.5% was practically trained, 31.1% only finished high school, and 1.7% did not have a diploma). Measurements. The measurements in Experiment 3 do not differ from those in Experiment 2. However, in Experiment 3, no additional explorative variables were measured because the main reason for performing this experiment was replication. Therefore, the measured variables were limited to moral justification (explained variance = 91.78%, Cronbach’s alpha = .91, M = 4.2, SD = 1.7), deservingness justification (explained variance = 93.56%, Cronbach’s alpha = .93, M = 3.1, SD = 1.8), impulse buying urge (explained variance = 91.34%, Cronbach’s alpha = .95, M = 3.0 , SD = 1.8), impulse buying behavior (one item), who the participants wanted to benefit from the purchase (one item; ranging from 1 = only to benefit myself to 7 = only to benefit others, M = 3.6, SD = 1.9), and the demographic variables sex, education, and age.
RkJQdWJsaXNoZXIy MjY0ODMw