The purpose of this study was to replicate the finding that not all biological explanations for crimes are equally mitigating. Specifically, while even mild traumatic brain injury (TBI) presents mitigating conditions for crimes committed during self-control failure, having an equal level of self-control (and failure) due to genetic reasons does not provide such mitigation. Participants read a story about someone who, in the heat of an argument, failed to control themselves and attacked another person, permanently blinding them in one eye. The cause for this low level of self-control was due to either genes or mild traumatic brain injury (TBI; randomly assigned). In our first study, we previously showed that while TBI provides mitigating circumstances leading to milder punishment, genes for low self-control do not create such a mitigating factor (compared to a control condition). Here, we sought to replicate this critical difference and explore its possible reasons. The primary dependent variable was whether participants would be less likely to find someone guilty of assault and battery after one attacked and permanently injured someone else. Participants were also randomly assigned to take the study as either part of the first set of 750 or the second set of 750 participants. We analyzed the 1st 750 first, followed by the 2nd 750, and then the two datasets combined. First 750 For our primary DV, we used a probit regression on whether participants would find the attacker guilty. We found that, in the first half of the data, participants were less likely to find the actor guilty if his low self-control was caused by a brain tumor (M = 85.042%) than if it was caused by his genes (M = 92.288%, bprobit (748) = -.386, 95%CI = -.144 to -.628, d = -.232, 95%CI = -.088 to -.375). To put the numbers in context: given a jury trial, this would correspond to an additional jury member voting not guilty. Second 750 This result was replicated in the second group of 750 participants, people were less likely to find the attacker guilty when his self-control failure was caused by TBI (M = 83.938%) than if caused by genes (M = 93.681%, bprobit (748) = -.537, 95%CI = -.285 to -.788, d = -.352, 95%CI = -.195 to -.508). Again, this would correspond to another juror voting not guilty in a hypothetical trial. Full 1500 These two effects represented a significant overall effect as well in the full data with participants being less likely to pronounce the man guilty if his self-control failure was caused by TBI (M = 84.471%) than if it was caused by genes (M = 92.961%, bprobit (1498) = -.459, 95%CI = -.285 to -.633, d = -.27, 95%CI = -.169 to -.372). There was no difference in effect size between the first and second 750 participants (bprobit (1496) = .15, p > .39). Secondary analyses We next tested different hypotheses about why traumatic brain injury represents a mitigating factor in crimes committed under self-control failure while being born with genes for an equivalent level do not. Due to the number of dependent variables taken, we took steps to protect against inflation of type I errors. With multiple dependent variables, the best ways to control for type I error inflation from multiple testing is to model all of the data concurrently in an SEM framework or incorporate Bonferroni-correction into each individual analysis (e.g. Bird & Hadzi-Pavlovic, 2014). Both strategies lead to the same conclusions in all but one instance, which we discuss. All results presented come from the structural model including all dependent variables simultaneously. We used robust maximum likelihood estimation to accommodate the use of Likert scales (Li, 2016). Level of self-control, having something ‘wrong’ with them We tested whether participants thought the attacker exerted different levels of self-control before attacking, and whether participants felt there was something ‘wrong’ with the attacker. Participants were asked: How much self-control did J--- exert before punching the man and Based on J-- having [genes/minor brain damage] causing below-average self-control, to what extent would you say that there is something ‘wrong’ with J---? The difference in perceived self-control between TBI and genes conditions was small (β = .06) and did not withstand multiple-testing correction. In addition, there was no difference in the extent to which participants perceived something was ‘wrong’ with the actor as a function of having either genes or mild TBI causing self-control failure (β = .024, p > .34). Thus, either of these two factors do not drive our results. Additional punishment Aside from whether participants would find the man guilty, we also asked, assuming he was found guilty, whether the attacker should be sent to prison or put on probation. If the biological cause was TBI, participants were less likely to recommend prison (M = 32.53%) than if the cause was genes (M = 51.527%, blogit = -.791, p < .001, 95%CI = -.581 to -1). Thus, the harshness of punishment was also reduced in self-control failure from TBI but not from genes. Fault and Blame Although in both instances the actor was the cause of the victim going permanently blind in one eye, we observed different levels of fault and blame attached to the attacker. Participants were asked How much at fault is J--- for the man going blind in one eye and How much should the victim blame J--- for making him go blind in one eye? We found both ratings of fault and blame were mitigated in the TBI versus the genes condition. Participants believed the victim should blame the attacker less in the TBI condition (M = 3.763) than in the genes condition (M = 4.068, β = -.141, p < .001, 95%CI = -.091 to -.192). Furthermore, they believed the attacker was less at fault if his self-control failure was caused by TBI (M =3.846) than if his self-control failure was caused by genes (M = 4.191, β = -.167, p < .001, 95%CI = -.117 to -.216). True self Finally, we tested whether participants believed the different biological causes reflected differences in who the ‘true self’ of the attacker was. Participants were asked how much they agreed with the statement Deep down, the true self of J--- is not a violent person. We argue the more participants agreed with this statement reflects the idea that their violent behavior was not attributable to their ‘true self’ but instead attributed to the biological cause. We found that people agreed more with this statement in the TBI condition (M = 4.079) than in the genes condition (M = 3.397, β = .225. p < .001, 95%CI = .177 to .274). We can see that the most likely reason for the different levels of mitigation was people attributed the violent act to the TBI and not to the agent’s ‘true self’. The same attacker with equal levels of self-control (and failure) caused by genes was held more responsible, more at fault, more blameworthy, and more deserving of harsher punishment, because their ‘true self’ was the one doing the attacking. New Variable: Having a condition Participants were finally asked: “Based on J-- having [genes/ minor brain damage] causing below-average self-control, to what extent would you say that this person has a condition? People thought someone who had low self-control failure as a result of minor TBI ‘had a condition’ to a greater extent (M = 3.356) than if they had genes for low self-control (M = 3.088, β = .117, p < .001, 95%CI = .067 to .167). Given that people differentially believed that the actor had a condition based on whether their low self-control was caused by TBI or having genes for low self-control, we tested an exploratory indirect effect analysis through ‘having a condition’ to each of the five significant effects (whether guilty, sending to prison vs probation, guilt, blame, and ‘true self’ beliefs. In four of the five cases case, people’s beliefs about the actor ‘having a condition’ was a partial indirect effect of the resulting beliefs (all ps < .037), with residual direct effects (all ps < .001). Furthermore, there was no indirect effect on beliefs in a non-violent ‘true-self’ (total indirect effect β = -.002, p > .12; residual direct effect β = .228, p < .001). Thus, while ‘having a condition’ may be seen as a partial explanation for mitigation, it is not so for beliefs about one’s ‘true self’.