Asymmetric Genetic Attributions for One’s Own Prosocial Versus Antisocial Behavior

Abstract

People tend to rate prosocial or positive behavior as more strongly influenced by the actor’s genes than antisocial or negative behavior. The current study tested whether people would show a similar asymmetry when rating the role of genes in their own behavior, and if so, what variables might mediate this difference. Participants were prompted to think about an example of their own behavior from the past year that was either prosocial or antisocial. Those in the prosocial condition rated the role of genetics in causing the behavior as significantly greater than did those in the antisocial condition. A mediation analysis suggested that this asymmetry could be accounted for by a tendency to view prosocial behavior as more natural and more aligned with one’s true self than antisocial behavior. These findings add to a growing body of evidence suggesting that people’s reasoning about genetics may be influenced by evaluative judgments.

The so-called “first law of behavior genetics” states that “All human behavioral traits are heritable” (Turkheimer, 2000, p. 160). Since its articulation, even more evidence has accumulated that genetic differences contribute to population-wide variance in behavioral traits, from food preferences and physical activity to antisocial and prosocial behavior (Conway & Slavich, 2017; de Geus, Bartels, Kaprio, Lightfoot, & Thomis, 2014; Ferguson, 2010; Fildes et al., 2014). Although people certainly do not always view genes as the primary determinant of human characteristics and behaviors—especially those that are seen as primarily psychological or mind-based, rather than body-based (Condit, 2011; Willoughby et al., 2019)—people appear to take a keen interest in understanding the influence of genes on various aspects of their lives. This is clear from the ever-increasing market for direct-to-consumer genetic testing services (National Academies of Sciences Engineering & Medicine, 2020) and the extent to which genetic findings generate media attention (Bubela & Caulfield, 2004). Gene-related metaphors have become part of everyday communication (Dar-Nimrod & Heine, 2011; Heine, 2017), and phrases like “it’s in my genes” have become commonplace.

Recent research has suggested that subjective evaluations of traits and behaviors may influence the extent to which laypeople attribute them to genetic causes. For example, when reading about a person behaving prosocially or antisocially, people tend to rate the prosocial behavior as more genetically influenced than the antisocial behavior (Lebowitz, Tabb, & Appelbaum, 2019). Similar asymmetries have been found for other traits; i.e., physical attractiveness is rated as more genetically influenced than physical unattractiveness, and being organized is rated as more genetically influenced than being disorganized (Lebowitz, Tabb, & Appelbaum, 2021). It has been posited that weaker genetic attributions for antisocial (versus prosocial) behavior may stem from a desire to view antisocial behavior as the product of free will, to justify holding wrongdoers responsible for their misdeeds (Lebowitz et al., 2019). That is, people might be less willing to entertain genetic attributions for antisocial behavior because genetic influences may be viewed deterministically, undercutting the ability to hold transgressors responsible for choosing their actions. Such an account could exemplify “blame validation processing,” which describes a “proclivity to favour blame versus nonblame explanations for harmful events” (Alicke, 2000, p. 568). Moving beyond this blame validation account, however, recent work has suggested that a more powerful mediator of the asymmetry in genetic attributions between positively and negatively valenced phenotypes may be perceptions of naturalness (Lebowitz et al., 2021), which are strongly linked to genetic attributions (Dar-Nimrod & Heine, 2011; Lynch, Morandini, Dar-Nimrod, & Griffiths, 2019).

While it has been established that people tend to view others’ positive traits and behaviors as more genetically influenced than their negative traits and behaviors, it remains unknown whether self-perceptions are subject to a similar asymmetry in genetic attributions. On one hand, to the extent that weaker genetic attributions for antisocial (versus prosocial) behavior are driven by blame validation processing, this kind of asymmetry might be unlikely to occur because people are often disinclined to favor blame-based explanations of their own misdeeds (Alicke & Sedikides, 2009; Snyder, Stephan, & Rosenfield, 2018). On the other hand, people tend to favor “internal” attributions for their own positive behaviors (Malle, 2006); they also generally rate their own “true selves” as good (Strohminger, Knobe, & Newman, 2017; Zhang & Alicke, 2021) and often view genes as akin to a person’s “essence” (Dar-Nimrod & Heine, 2011). As such, people might be motivated to attribute their own positive behaviors and traits, but not their negative ones, to their genes, to preserve the view of their essential “true selves” as good. Although previous research did not find that asymmetries in genetic attributions for others’ prosocial and antisocial behaviors were mediated by asymmetrical ascriptions of behavior to the “true self” (Lebowitz et al., 2019), motivations to view the “true self” in a positive light might be stronger for oneself than for others (Zhang & Alicke, 2021).

The present study examined whether people would make stronger genetic attributions for their own prosocial behavior than antisocial behavior, and if so, whether any of the three mediators discussed above—ascriptions of responsibility, “true self” judgments, and naturalness ratings—would mediate such an asymmetry.

Materials and Methods

Participants.

Participants were 600 U.S. adults (54.0% female, 43.5% male, 2.5% other or unknown gender; mean age 32.87 years, SD=11.05) recruited from the online platform Prolific, which allows participants to earn money in exchange for completing online experiments (Palan & Schitter, 2018).

Procedures.

After participants provided informed consent via an online form, they were randomly assigned to either a prosocial condition (n=301) or an antisocial condition (n=299). Then, the following prompt was displayed: “Please take a moment to think of one example of your own behavior from the past year that you are most [ashamed/proud] of. For example, you might think about the most [selfish/generous] or [harmful/helpful] thing that you can remember doing in the past year.” For the bracketed terms, the word before the slash was displayed to participants in the antisocial condition, whereas the word after the slash was displayed to participants in the prosocial condition. The “past year” timeframe was selected to be long enough that participants would have a range of behaviors from which to choose, but short enough that participants would be relatively likely to remember their behaviors. After the prompt, participants were instructed: “Keeping in mind this thing that you remember doing, please answer the following questions.” They were then asked to provide a naturalness rating, a responsibility rating, and a true-self rating, all of which were examined as possible mediators of prosocial/antisocial differences in genetic attributions and were measured in the same way as in prior research (Lebowitz et al., 2019, 2021). The three measures were displayed in a randomized order.

Naturalness ratings were provided in response to the question, “How natural was it for you to do this thing that you did?” with responses on a scale from “1 (Not at all natural)” to “7 (Very natural).” Responsibility ratings were provided in response to the question, “To what extent were you responsible for doing this thing that you did?” with responses on a scale from “1 (Not at all responsible)” to “7 (Very responsible).” True-self ratings were provided in response to the question, “To what extent did this thing that you did reflect your true self — the person you truly are, deep down?” with responses on a scale from “1 (Not at all)” to “7 (Very much).” Finally, participants provided a genetic attribution rating in response to the question, “How much of a role did your genetics play in causing you to do this thing that you did?” on a scale from “1 (No role or a very minor role)” to “7 (A very major role).”

At the end of the study procedures, participants were asked demographic questions and viewed a basic debriefing. See https://osf.io/2t7ks/files/ for data used for analysis, matrices of correlations among measures, and the data-collection survey.

Results

We first used a one-way ANOVA to examine the effects of condition (prosocial vs. antisocial) on naturalness ratings, responsibility ratings, true-self ratings, and genetic attribution ratings (see Figure 1) and caculated Cohen’s d for each prosocial-antisocial mean difference. Naturalness ratings were significantly higher in the prosocial condition (M=5.31, SD=1.45) than in the antisocial condition (M=3.67, SD=1.64), F(1, 598)=167.33, p<.001, d=1.06. Responsibility ratings were not significantly different between the prosocial (M=6.08, SD=1.17) and antisocial (M=6.04, SD=1.24) conditions, F(1, 598)=.16, p=.687, d=.03. True-self ratings were significantly higher in the prosocial condition (M=5.79, SD=1.19) than in the antisocial condition (M=3.40, SD=1.79), F(1, 598)=372.37, p<.001, d=1.57. Genetic attributions ratings were also significantly higher in the prosocial (M=3.25, SD=1.92) than antisocial (M=2.47, SD=1.72) condition, F(1, 598)=27.44, p<.001, d=.43.

Figure 1.

Mean ratings, by condition, for measured variables. Error bars represent +/− 1 standard error.

To test naturalness, responsibility, and true-self ratings as potential mediators of the prosocial-antisocial difference in genetic attributions, we used the PROCESS (Hayes, 2018) procedure (version 3.4) for SPSS with 5,000 bootstrap samples. For this analysis, we used PROCESS model 4, designating condition (prosocial coded as 1, antisocial coded as 0) as the independent variable (X) and genetic attribution ratings as the dependent variable (Y); naturalness ratings, responsibility ratings, and true-self ratings were all entered in parallel as mediator (M) variables. The results are shown in Figure 2. Of note, there were significant indirect effects of condition on genetic attributions through naturalness ratings, B = .39, SE = .10, 95% percentile bootstrap CI (.20, .61), and through true-self ratings, B = .43, SE = .15, 95% percentile bootstrap CI (.13, .73); the indirect effect through responsibility ratings was not significant, B = −.001, SE = .01, 95% percentile bootstrap CI (−.02, .01).

Figure 2.

Analysis of naturalness ratings, responsibility ratings, and true-self ratings as potential mediators of the effect of prosocial (versus antisocial) behavior on genetic attribution ratings.

Discussion

The present results suggest that people make stronger genetic attributions for their own prosocial behavior than for their antisocial behavior. Although previous research had indicated that similar asymmetries occur when people make genetic attributions for behaviors and traits described in others (Lebowitz et al., 2019, 2021), the current research is the first to extend this finding to situations in which people are reflecting on their own actions. Moreover, the present study examined three possible mediators for such an asymmetry in genetic attributions: ascriptions of responsibility, “true-self” judgments, and naturalness ratings. Unlike previous research examining the genetic attributions that participants made after reading descriptions of others’ behaviors (Lebowitz et al., 2019), the current study found that asymmetries in the extent to which the two types of behavior were seen as reflective of the “true self” were a significant mediator of the asymmetry in genetic attributions, but ascriptions of responsibility for the behavior were not. However, the present study mirrored other past research (Lebowitz et al., 2021) in suggesting that the perception of positive behavior as more “natural” than negative behavior may help to account for the asymmetry in genetic attributions.

One question raised by this study’s results is whether different mediators actually account for asymmetrical genetic attributions for prosocial and antisocial behavior, depending whether one is explaining one’s own behavior or for another person’s. For example, does differential willingness to attribute positive (versus negative) behaviors to the “true self” play a role in asymmetrical genetic attributions for one’s own behavior, but not for that of others? The present study cannot directly answer this question, having examined ratings only of participants’ own behavior. However, it is plausible that concerns about maintaining support for the notion of an inherently good true self may operate more strongly in the context of self-reflection than in situations where one is judging others. For instance, people tend to judge immoral behaviors as less characteristic and moral behaviors as more characteristic of their own true self, as compared to another person’s true self (Zhang & Alicke, 2021). Future research could further clarify why people view genetic attributions as more plausible in the case of positively valenced behavior than negatively valenced behavior, as well as the extent to which the answer might differ depending whether the behavior being judged is one’s own or another’s. Regardless, it seems clear that perceptions of positive behavior and characteristics as more natural than negative behaviors and characteristics (Lebowitz et al., 2021) are a robust mediator of asymmetries in genetic attributions, whether the the target being considered is oneself or another person.

The present study is not without limitations. For one, because we wished participants to consider their own behavior, it was not possible to use standardized examples of prosocial and antisocial behavior that could be held constant across participants. Moreover, the instructions prompted participants to think of a behavior they exhibited in the past year of which they were “ashamed” (antisocial condition) or “proud” (prosocial condition). Although they were given examples intended to evoke memories of prosocial behaviors (“you might think about the most generous or helpful thing that you can remember doing”) or antisocial behaviors (“the most selfish or harmful thing”) depending on their assigned condition, it is possible that not all participants were truly reflecting on prosocial or antisocial behavior. More generally, there are likely wide individual differences in the types of behavior that participants were drawing on, although it is important to note that this is unlikely to confound the main effects of condition observed in the present study, given the use of random assignment.

Additionally, prompting participants to consider behaviors of which they were “proud” or “ashamed” may have ensured that they would recall behaviors for which they had already taken responsibility, which could help to explain the lack of a difference in responsibility ratings between the conditions. It is also possible, especially in the antisocial condition, that genetic attributions would be higher (and responsibility ratings potentially lower) in a context where participants were more motivated to disavow responsibility for their actions (which may not have been the case in our online experiment). Nonetheless, an asymmetry in genetic attributions was observed despite no asymmetry in responsibility ratings, consistent with the interpretation that the former is not dependent on the latter, although it is not inconceivable that the asymmetry in genetic attributions could have been stronger if there had been an asymmetry in responsibility ratings.

Also, the “proudest” and “most ashamed” prompts may have elicited consideration of particularly extreme behaviors; it is unclear whether the same pattern would be observed in response to less extreme examples. However, past research investigating reactions to the behavior of others, rather than one’s own, suggests that asymmetries in genetic attributions occur across a wide range of prosocial and antisocial behaviors (Lebowitz et al., 2019).

Furthermore, because the present study compared ratings of one’s own prosocial versus antisocial behaviors, the findings can only illuminate differences in ratings between these two types of behaviors, not how the prosocial or antisocial nature of a behavior affects genetic attributions (or its mediators) in a more absolute sense. Future research could examine this question by comparing ratings of prosocial and antisocial behaviors against those of morally neutral ones.

The present study extends the prior finding that positively valenced behaviors are seen as more genetically influenced than negatively valenced ones, and demonstrates for the first time that this asymmetry occurs when people are considering their own behaviors. This finding has implications for a range of domains of reasoning about the self and identity, as it suggests that people view their own positive behaviors as reflective of their “true selves” and even their DNA, whereas this is not the case (to the same extent) for their negative behaviors. This may help to illuminate a psychological mechanism through which people employ motivated reasoning to build and preserve a positive self-concept. It may also have implications for science literacy and even health, if people are more willing to believe genetic explanations for their positively valenced qualities (e.g., socially desirable attributes or positive health states) than for their negative ones, even in the face of equivalent evidence for both. Although the “first law of behavior genetics” may tell us that all human behavior is heritable, the present findings add to a growing body of evidence suggesting that people may be selective about the kinds of behavior that they are inclined to attribute to genetic causes.