What are you calling intuitive? Subject heterogeneity as a driver of response times in an impunity game

Abstract

We study choices and reaction times of respondents in an impunity game with unfair offers. The non-private impunity game features two roles, proposer and respondent, who are both aware whether the pie size is small or large. Proposers decide among three more or less unfair offers; respondents can accept or reject the offer, in which case it is lost for them. Whatever the responder decides is communicated to the proposer.240 proposers took part in a traditional laboratory; 24 respondents were in an fMRI setup where they confronted all 240 proposals elicited from proposers. Responses were sent via email to proposers.Proposers revealed little concern for respondents. Respondents overwhelmingly rejected small offers, especially from a large pie. Surprisingly and in contrast with most of the literature and the Social Heuristic Hypothesis, we find that on average acceptances took longer than rejections. This result is driven by individual heterogeneity. The rich response data allow us to distinguish different respondent types, finding a remarkable consistency: subjects mainly accepting (rejecting) take more time to reject (accept). We attribute this finding to heterogeneity in self-priming. Our results suggest a primary role for individual heterogeneity in experiments testing the intuitive or deliberate status of cooperation and altruism.

Delaying and Motivating Decisions in the (Bully) Dictator Game

“…In this section, we attempt to provide answers to the research questions by relying on the method of analysis that we pre-registered alongside an a priori power analysis for our sample size. Given the novelties introduced with our treatments, we took a conservative approach and posited an effect size d = 0.13 -defined as the ratio between population means and the overall population standard deviation (Cohen, 2013) -in the range of what is considered small in the literature (Cohen, 2013;Lakens, 2013;Thompson, 2007). We then computed our sample size by setting an expected statistical power of 0.8 for a significance level of 5%, which required 164 subjects per treatment (rounded up to 170).…”

“…Some conditions use time-related manipulations, others rely on priming or involve tasks that require cognitive effort. However, it is unclear to what extent the impact on behavior is sensible to the specific features of these conditions, i.e., the specific mechanism through which they affect prosocial behavior (Chen & Krajbich, 2018;Crosetto & Güth, 2021). Regarding the conditions aimed at promoting reliance on intuition, although there is no systematic comparative study on their effects, some evidence has been accumulated which suggests that relying on time pressure (Alós-Ferrer & Garagnani, 2020a;Merkel & Lohse, 2019;Teoh, Yao, Cunningham, & Hutcherson, 2020), cognitive load (Achtziger, Alós-Ferrer, & Ritschel, 2020;Deck & Jahedi, 2015;Schulz, Fischbacher, Thoni, & does not.…”