| Abstract: | Survey satisficing in online data collection biases the estimation of treatment effects in many ways. Extending the findings of a previous study, which demonstrated that satisficing biased the estimation of main treatment effects, this study also shows that satisficing distorts the estimation of moderation effects. Targeting Japanese adults’ attitudes toward food, this study tests how the effect of country of production (Japan vs. China) is moderated by preexisting ethnocentric attitudes. The results show that while nonsatisficers predictably adjust their attitude toward food based on their preexisting ethnocentric attitude, satisficers stick to their initial stereotypical response. That is, the theoretically predicted moderation effect was observed among nonsatisficers, but not among satisficers, which indicates that satisficing biases not only the estimation of a treatment effect but also that of a moderation effect. |
