34 The Urge to Splurge In two out of three experiments we found an indirect effect of interactivity on impulse-visit urges through self-agency. In Experiment 1, we only found an indirect effect of interactivity through self-agency on impulse-buying urges, and not on impulse-visit urges. As discussed previously, the lack of this effect could be attributed to the smaller sample size of Experiment 1 (compared to Experiments 2a and 2b). An alternative explanation could be that the explaining mechanism in the effect of interactivity on impulse visit urges partially differs from the explaining mechanism in the effect of interactivity on impulse buying urges. So far, the literature did not give reasons to suspect that the effect of interactivity on both dependent variables would be explained by different mediators (see e.g., Prediger et al., 2019). However, the results of Experiment 1 indicate that it would be interesting to explore the existence of different mediators in the effect of interactivity on impulse visits and impulse purchases. In line with our expectations, all three field experiments showed that self-agency functions as a mediator in interactivity’s effect on impulse-buying urges. This corresponds with the AMC, which suggests that interactivity techniques could enhance agency and that self-agency could affect attitudes toward displayed items (Sundar 2008). By proving that self-agency can enhance impulse urges, we introduce the term ‘the self-agency paradox,’ which means that when people think they are in charge of what happens and feel as though they cause certain events (self-agency), they are actually less in charge than they assume (Kang and Sundar, 2016) and are more likely to yield to their impulses. A possible explanation for this paradox can be found in Reactance Theory, which states that people who feel free to make their own choices (which is often intertwined with feelings of self-agency [Barlas and Obhi, 2013]) are easier to persuade. However, additional research must be conducted to substantiate the existence of the self-agency paradox. We did not find the proposed direct effect of interactivity on impulse-visit urge and impulse-buying urge. A possible explanation for this is that interactivity could both and simultaneously have a negative and positive effect on impulse urges. It appears that the residual direct effect of interactivity on impulse visit urges and impulse buying urges, when taking self-agency into account, turns out to be negative in two out of three of our experiments (Experiment 1buying: B = -.69, SE = .30, 95% CI: [-1.29 to -.09], p = .025; Experiment 2avisit : B = -.54, SE = .27, CI: [-1.07 to -.01], p = .046; Experiment 2abuying : B = -.62, SE = .23, 95% CI: [-1.06 to -.17], p = .007). This means that two processes could be triggered by interactivity that cancel each other out, whereof self-agency is the process that explains the positive effect of interactivity. Future studies should examine the possible mediators that could explain the negative effect of interactivity, such as involvement (see Liu & Shrum, 2013). Implications for Practice This study provides practical contributions to professionals working in the retail, marketing, and advertising fields. We gained more relevant and externally valid information for retailers that run a physical store by conducting field experiments. Hereby, this study explicitly distinguished impulse-visit urges from impulse-buying urges. We proposed that impulse visits are a manifestation of experiential shopping motives and behavior while impulse purchases would pertain to more materialistic motives and behavior. Since (impulsive) store visits do not necessarily lead
RkJQdWJsaXNoZXIy MjY0ODMw