Access the full text.
Sign up today, get DeepDyve free for 14 days.
Prior research has indicated that disease threat and disgust are associated with harsher moral condemnation. We investigated the role of a specific, highly salient health concern, namely the spread of the coronavirus, and associated COVID-19 disease, on moral disapproval. We hypothesized that individuals who report greater subjective worry about COVID-19 would be more sensitive to moral transgressions. Across three studies (N ¼ 913), conducted March-May 2020 as the pandemic started to unfold in the United States, we found that individuals who were worried about contracting the infectious disease made harsher moral judg- ments than those who were relatively less worried. This effect was not restricted to transgressions involving purity, but extended to transgressions involving harm, fairness, authority, and loyalty, and remained when controlling for political orientation. Furthermore, for Studies 1 and 2 the effect also was robust when taking into account the contamination subscale of the Disgust Scale–Revised. These findings add to the growing literature that concrete threats to health can play a role in abstract moral considerations, supporting the notion that judgments of wrongdoing are not based on rational thought alone. Keywords morality, disgust, pathogen avoidance, behavioral immune system, moral judgment, emotion, harm, COVID-19, coronavirus, moral foundations theory Date received: November 28, 2020. Revision Submitted: April 17, 2021; Accepted: May 13, 2021 People’s moral compass is typically assumed to be firmly be attributed to infections (Finch, 2010, 2012). Even in armed grounded in rational thought. For example, legal systems rely conflict, illness has historically accounted for far more deaths on judges and jurors making decisions about wrongdoing via than those that result from combat itself. For example, in the detached evaluation of the available evidence. However, an American civil war, two-thirds of the estimated 660,000 deaths emerging literature suggests that judging right or wrong can were caused by pneumonia, typhoid, dysentery, and malaria be colored by factors that are objectively unrelated to deliberate (Connolly & Heymann, 2002). Furthermore, during famines, considerations, such as emotions and intuitions. Haidt (2001) infectious diseases have caused more deaths than starvation as proposed that such factors are the driving force behind moral a consequence of the behavioral changes induced by conditions judgments, with rationalizations taking place only after a deci- of extreme hunger (Shaw-Taylor, 2020). Even in modern times, sion has already been reached. Indeed, there is an increasing nearly a quarter of all worldwide deaths have been due to recognition that morality is shaped by processes that unfold infectious diseases—more than double that from violence or largely outside of conscious awareness. injury (WHO, 2015). In particular, disgust has been suggested to play a role in the evaluation of moral transgressions due to its evolutionary func- Department of Psychology, University of Cambridge, United Kingdom tion of disease avoidance (Rozin et al., 2008; Schnall et al., 2008; Tybur et al., 2013). Indeed, pathogens and parasites have Corresponding Author: played an outsized role in evolutionary history. For both Robert K. Henderson, Department of Psychology, University of Cambridge, hunter-foraging societies and our nearest evolutionary rela- Downing Street, Cambridge CB2 3EB, United Kingdom. tives, chimpanzees, about seven out of every 10 deaths can Email: email@example.com Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Evolutionary Psychology In light of the substantial risks posed by infectious illnesses, the U.S. government (March 13, 2020). Occurrence was largely there has been a large body of literature investigating the asso- concentrated in Washington state, with 904 cases, including 48 ciations between disease threat, behavioral caution, lower tol- deaths. We therefore sampled participants from this state, and erance for nonconformity, and political conservatism as a comparison, Maine, a less densely populated state with (Mortensen et al., 2010; Murray & Schaller, 2012; Wu & only 17 cases and no deaths at that time. We reasoned that fear Chang, 2012; Zmigrod et al., 2020). Such associations are of the virus would be higher in the former than the latter state. thought to reflect the potency of disease threat, such that infec- In addition, to make the health threat salient, half the partici- tious disease concerns motivate individuals to more closely pants read a New York Times article on the dangers of the behave in line with societal expectations. pandemic while the other half read a neutral article about Likewise, disease threat appears to be associated with moral national parks. We predicted harsher moral judgments for par- vigilance (Murray et al., 2019; Park & Isherwood, 2011; Van ticipants who were worried about catching the virus, compared Leeuwen et al., 2012). Results from this line of research are to those who were not. consistent with the conceptual link between disgust and moral considerations, as disgust is thought to have evolved primarily to facilitate disease avoidance. Historically, individuals believed Method that violating moral proscriptions increased the likelihood of Participants danger, and in particular the spread of infectious disease (Fabrega, 1997). Therefore, wrongdoers who violated such Participants from Washington and Maine were recruited via the norms posed a threat to the survival of others. Under perilous online participant panel Prolific. Because it was the first study, conditions, such as during a pandemic, such norms may take on we did not have a specific effect size in mind, and aimed for a even more importance, especially to the extent that individuals target sample of 200, collecting data from 220 participants in subjectively evaluate the infectious disease as threatening. anticipation of possible exclusions. We removed data from 14 Indeed, Murray et al. (2019) found a positive association participants for failing attention checks. The final sample con- between sensitivity to moral wrongdoing and the germ aversion sisted of 206 participants (130 women; age: M ¼ 36.80 years, subscale of the Perceived Vulnerability to Disease scale. SD ¼ 14.16), with 165 from Washington, and 41 from Maine. Another way of approaching this issue is to examine natu- rally occurring concerns about physical contamination, such as Procedure the fear of contracting a highly salient contagious disease that poses an immediate threat. In other words, to explore the rela- After providing informed consent, participants were randomly tionship between disease threat and morality, one can examine assigned to read one of two New York Times articles, either on the relationship between concern about physical health and the dangers of coronavirus infections, or about national parks, moral judgments directly. In early 2020, the global spread of both published on March 13, 2020. They then responded to a previously unknown type of coronavirus (SARS-CoV-2) 60 Moral Foundations Vignettes that had been pre-tested and leading to COVID-19 disease presented such an opportunity. standardized (Clifford et al., 2015). There were 12 violations In March-May 2020 we assessed whether U.S. participants’ for each foundation, rated on a scale from 1 (not at all wrong) to fear about catching the disease was related to their moral judg- 5 (extremely wrong). Scenarios included, “You see a girl ments. We did so by asking a standard polling question about laughing when she realizes her friend’s dad is the janitor” coronavirus worry, and administering a set of survey items that (Harm), “You see a tenant bribing a landlord to be the first encompassed different domains of morality. Moral Founda- to get their apartment repainted” (Fairness), “You see a man tions Theory (Graham et al., 2009) proposes at least five moral leaving his family business to go work for their main foundations: Aversion for the suffering of others (Harm), con- competitor” (Loyalty), “You see a star player ignoring her cern with cheating and lack of reciprocity (Fairness), group coach’s order to come to the bench during a game” (Author- adherence (Loyalty), deference to leadership and tradition ity), and “You see two first cousins getting married to each (Authority), and concern with purity and contamination (Pur- other in an elaborate wedding” (Purity). Vignettes were admi- ity). These five foundations are thought to have arisen to cope nistered in a randomized order. with adaptive challenges in human ancestral environments. We Then participants indicated their worry about COVID-19 hypothesized that individuals who report subjective worry by responding to a standardized question taken from the pop- about contracting COVID-19 would express more disapproval ular public opinion and data company YouGov: “Taking into when evaluating wrongdoing than individuals with relatively consideration both your risk of contracting it and the serious- lower worry. ness of the illness, how worried are you personally about experiencing coronavirus?” Response options included “not at all worried,” “not too worried,” “somewhat worried” and Study 1 “very worried.” Participants then completed the Disgust The first study was conducted on March 17, 2020, days after Sensitivity Scale-Revised (Haidt et al., 1994, modified by COVID-19 was declared a pandemic by the World Health Olatunji et al. (2007)) to assess whether the contribution of Organization (March 11, 2020), and a national emergency by coronavirus worry went above and beyond this individual Henderson and Schnall 3 Figure 1. Ratings of moral condemnation for individuals who were worried, or relatively less worried about coronavirus, Study 1. difference variable. Lastly participants provided demo- who were more worried (i.e., indicated they were “somewhat graphics and their political orientation, rated on a scale from worried” or “very worried,” n ¼ 134). Furthermore, because 1(¼ very liberal) to 7 (¼ very conservative), and were the study was administered less than 1 week after COVID-19 debriefed and compensated. was declared a national emergency in the U.S., and most parti- cipants reported at least some level of worry about COVID-19, we dichotomized the variable into “less worried” and “worried.” Dichotomizing continuous variables can be a useful Results approach for analyzing non-normal data (Farringdon & Loeber, Manipulation Check 2000). Indeed, a Shapiro-Wilk test showed a significant depar- ture from normality, W(206) ¼ .84, p < .001. For each parti- Using a 2 2 ANOVA we first tested whether COVID-19 cipant, the mean moral disapproval rating across the five case prevalence as a function of state (Washington vs. foundations was calculated, with higher scores indicating more Maine) and Virus Threat Salience (Coronavirus vs. National severe condemnation. Parks article) were associated with different levels of worry We then performed a repeated-measures ANOVA with about contracting the illness. Unexpectedly, there was no Moral Foundation (Harm, Fairness, Authority, Loyalty, and effect of State, F(1, 202) ¼ .74, p ¼ .39, nor of Threat Purity) as a within-subjects factor and Worry (Worried vs. Salience, F(1, 202) ¼ .81, p ¼ .37, and no interaction, Less-Worried) as a between-subjects factor. The Huynh-Feldt F(1, 202) ¼ .03, p ¼ .87, indicating that for people’s con- correction was applied because Mauchly’s test of sphericity cern about the virus it did not matter as a function of was significant (p < .001). There was a main effect of moral whether they resided in an area with high vs. low disease foundation, F(3.47, 707.45) ¼ 123.43, p < .001, Z ¼ 0.38, prevalence, nor whether they had been primed with infor- with the highest ratings for purity (M ¼ 3.61, 95% CI ¼ [3.52, mation about the virus, or not. We therefore were unable to 3.70]) and fairness violations (M ¼ 3.48, 95% CI ¼ [3.40, analyze moral judgments based on these variables, and 3.55]), followed by harm (M ¼ 3.45, 95% CI ¼ [3.36, 3.53]), instead performed exploratory analyses using participants’ authority (M ¼ 3.17, 95% CI ¼ [3.08, 3.26]), and loyalty viola- self-reported level of coronavirus concern. tions (M ¼ 2.78, 95% CI ¼ [2.68, 2.89]). Testing the key prediction, worried participants (M ¼ 3.42, Moral Judgment SD ¼ .51) produced harsher moral judgments than less-worried Following standard practice in opinion polls from which we participants (M ¼ 3.18, SD ¼ .51), F(1, 204) ¼ 10.66, p ¼ .001, derived the survey item (YouGov), we divided the sample into Z ¼ .05 (see Figure 1 for means). Foundation Type did not participants who were less worried (i.e., indicated they were interact with Worry, F(3.55, 724.70) ¼ .49, p ¼ .722, suggest- “not at all worried” or “not too worried,” n ¼ 72), and those ing that the effect was comparable across foundations. 4 Evolutionary Psychology threats, in case this was important for subjectively experienced Political Orientation worry about the virus. Because prior research has shown that political orientation correlates with moral judgments (e.g., Graham et al., 2009), we also conducted an ANCOVA for which we added responses Method to this item as a covariate to the analysis above. Consistent with Participants earlier research, political orientation was a significant predictor of moral condemnation, F(1, 203) ¼ 15.63, p < .001. More We recruited participants via Amazon Mechanical Turk from importantly, however, when controlling for it, the effect Washington and Maine. Building on the observed effect size of coronavirus worry remained robust, F(1, 203) ¼ 14.97, from Study 1, d ¼ .47, a G*Power analysis (Faul et al., 2007) p < .001, Z ¼ .069. Thus, the observed difference was not indicated a required sample size of 238 for an independent driven by political ideology. samples t-test (two-tailed) with 95% power at a ¼ .05. We removed data from 11 participants because they did not com- plete the study, and from seven participants for failing attention Contamination Disgust checks. The final sample consisted of 220 participants (126 Across participants, scores on the contamination subscale of women; age: M ¼ 39.11 years, SD ¼ 12.69), with 189 from the Disgust Scale–Revised (DS-R, Haidt et al., 1994, modified Washington, and 44 from Maine. by Olatunji et al., 2007) was associated with COVID-19 worry (r ¼ .20, p < .001). To investigate the possible effects of con- Procedure tamination disgust on moral judgment between the two groups, a repeated-measures ANCOVA was performed with contami- Identical materials, measures and procedure as in Study 1 were nation disgust as a covariate. There was an effect for this cov- used, but there was no coronavirus manipulation at the begin- ariate, F(1, 203) ¼ 22.10, p < .001, Z ¼ .10. More ning, and administration of the moral stimuli and the corona- importantly, however, the main effect of coronavirus worry virus worry question was counterbalanced. remained significant after controlling for contamination dis- gust, F(1, 203) ¼ 6.28, p ¼ .013, Z ¼ .03. Results Discussion Manipulation Check Study 1 provided initial evidence that the extent to which peo- We first conducted a one-way ANOVA to test whether coro- ple were worried about coronavirus in the early days of the navirus worry differed between residents of Washington and 2020 COVID-19 pandemic related to condemnation of moral Maine. Consistent with the results from Study 1, there was no transgressions that were unrelated to the virus: Participants effect, F(1, 218) ¼ .002, p ¼ .97. We therefore used the same who were worried about COVID-19 produced harsher moral analysis strategy and focused on participants’ self-reported judgments than those who were less worried. Importantly, this level of coronavirus concern (i.e., their subjectively experi- effect could not be attributed to political orientation, consistent enced threat). with earlier findings that the role of disgust in moral judgment is not explained by ideology (van Leeuven et al., 2017). We Moral Judgment furthermore controlled for contamination disgust, and although We performed a repeated-measures ANOVA with Moral Foun- it accounted for some of the variance, the association between dations (Harm, Fairness, Authority, Loyalty, and Purity) as a worry and moral judgment remained robust. within-subjects factor and Worry (Worried vs. Less-Worried) However, we had included a threat salience manipulation at and Order (Worry Question first vs. Moral Judgments first) as the beginning of the study, which turned out to be ineffective, between-subjects factors. The Huynh-Feldt correction was used and considered it important to replicate the effect without such because Mauchly’s test of sphericity was significant (p ¼ .001). a procedure before participants made the moral judgments. We Results revealed no main effects of order F(1,216) ¼ 0.12, therefore conducted another study, and as additional improve- p ¼ .725, nor any Order Group interaction, F(1, 216) ¼ ment also counterbalanced the order in which the moral judg- 0.22, p ¼ .637. Therefore, order was not further considered. ments and the coronavirus worry question were administered. There was a main effect of moral foundation, F(3.48, 757.52) ¼ 75.602, p < .001, Z ¼ 0.26, with the highest ratings Study 2 for purity violations (M ¼ 3.65, 95% CI ¼ [3.55, 3.75]), then The second data collection was carried out 10 days later, on fairness (M ¼ 3.46, 95% CI ¼ [3.38, 3.55]), followed by harm March 27, 2020, when COVID-19 cases across the U.S. had (M ¼ 3.29, 95% CI ¼ [3.19, 3.39]), authority (M ¼ 3.20, 95% risen somewhat but were still relatively localized, with 3,700 CI ¼ [3.09, 3.31]), and loyalty violations (M ¼ 2.92, 95% CI ¼ cases in Washington state, including 174 deaths, but only 168 [2.80, 3.05]). Replicating the results of Study 1, there was a cases and one death in Maine, respectively. We again focused main effect of worry, such that worried participants (M ¼ 3.43, on these two states as representing objectively different virus SD ¼ .58) showed greater moral condemnation than less- Henderson and Schnall 5 Figure 2. Ratings of moral condemnation for individuals who were worried, or relatively less worried about coronavirus, Study 2. worried participants (M ¼ 3.18, SD ¼ .60), F(1, 218) ¼ 8.67, p Discussion ¼ .004, Z ¼ .04 (see Figure 2). There was no foundation This study replicated the observation that a situational threat to worry interaction, F(3.48, 757.52) ¼ .52, p ¼ .694, again indi- one’s physical health, namely concern about contracting an cating that the effect was not limited to any specific founda- illness that was spreading rapidly throughout the U.S. at the tions, such as purity. time, was related to moral concerns. Speaking to ongoing debates of whether the link between disgust and morality is domain-specific, or more general (see Schnall, 2017, for a Political Orientation discussion), in both studies the effect was not specific to trans- To rule out this possible confound, the same analysis as above gressions involving purity, but extended to all moral was conducted with political orientation as a covariate. Con- foundations. sistent with earlier research, political orientation was a signif- icant predictor of moral condemnation, F(1, 217) ¼ 29.64, p < .001. As was the case for Study 1, there still was a significant Study 3 difference between the worried and less-worried participants for moral disapproval, F(1, 217) ¼ 13.47, p < .001, Z ¼ .058, Both studies 1 and 2 included samples from only two areas of again showing the independent contribution of disease concern. the U.S. that had varying levels of cases of COVID-19. Parti- cipants in these states did not differ in subjectively perceived worry about the virus, thus justifying the use of the latter as the Contamination Disgust predictor variable. These findings still raise the question, how- Consistent with Study 1, contamination disgust was associated ever, of whether the same effect would be observed more with COVID-19 worry (r ¼ .20, p ¼ .003). To again investigate broadly across the population. In particular, while in mid-to- the possible effects of contamination disgust on moral judg- late March 2020, when the first two studies were conducted, ment between the two groups, a repeated-measures ANCOVA COVID-19 cases were relatively low in the U.S., it was also was performed with scores on the contamination subscale of important to explore whether the effects would persist as the the Disgust Scale-Revised (Olatunji et al., 2007) as a covariate. pandemic unfolded across the country. We therefore conducted Contamination disgust showed a significant effect regarding a preregistered replication about 6 weeks after Study 2, sam- moral condemnation, F(1, 217) ¼ 54.10, p < .001, Z ¼ .20. pling across the entire U.S. On May 6, 2020, when the study Importantly, and replicating the results from Study 1, the effect was conducted, there were 1,261,354 COVID-19 cases, includ- of coronavirus worry on moral judgment remained significant ing 74,710 deaths. We again predicted that people worried after controlling for contamination disgust, F(1, 217) ¼ 4.22, about the virus would rate moral infractions as more objection- p ¼ .041, Z ¼ .02. able than those who were less worried. p 6 Evolutionary Psychology Figure 3. Ratings of moral condemnation for individuals who were worried, or relatively less worried about coronavirus, Study 3. Replicating the earlier results, there was a main effect of Method moral foundation, F(3.37, 1631.87) ¼ 186.20, p < .001, Z ¼ Participants 0.28, with the highest ratings for Purity violations (M ¼ 3.70, Participants were recruited across the U.S. with Prolific. Using 95% CI ¼ [3.63, 3.76]), followed by fairness (M ¼ 3.49, 95% the effect size from Study 2, d ¼ .42, a G*Power analysis using CI ¼ [3.43, 3.56]), harm (M ¼ 3.45, 95% CI ¼ [3.39, 3.52]), 95% power at a ¼ .05 specified a required sample of 296 authority (M ¼ 3.26, 95% CI ¼ [3.19, 3.33]), and loyalty viola- participants. However, because the earlier studies were con- tions (M ¼ 2.88, 95% CI ¼ [2.80, 2.96]). ducted only in Maine and Washington, to account for increased Consistent with the findings from Studies 1 and 2, and as variability when sampling across the entire population of the specified in our preregistration (https://aspredicted.org/blind. U.S., we set our preregistered sample to 500. Data from 13 php?x¼2as8ic), worried participants (M ¼ 3.42, SD ¼ .59) participants were excluded because they failed attention exhibited harsher moral judgments than less-worried partici- checks. The final sample involved 487 participants (273 pants (M ¼ 3.30, SD ¼ .52), F(1, 485) ¼ 4.43, p ¼ .036, Z women; age: M ¼ 31.25, SD ¼ 11.77). ¼ .01 (see Figure 3). There was no Foundation Worry inter- action, F(3.37,1631.87) ¼ .53, p ¼ .686. Thus, as was the case for Studies 1 and 2, the specific moral foundation did not Procedure moderate the effect of coronavirus worry on moral judgments. The method was identical to Study 2. Political Orientation Results To once again test for possible effects of political orientation on moral judgment between the two groups, a repeated-measures We performed a repeated-measures ANOVA with Moral Foun- ANCOVA was performed as before. As with Studies 1 and 2, dations (Harm, Fairness, Authority, Loyalty, and Purity) as a political orientation had a significant effect on moral condem- within-subjects factor and Worry (Worried vs. Less-Worried) nation, F(1,484) ¼ 29.50, p < .001. Also replicating the earlier and Order (coronavirus question first or moral foundations findings, there was a statistically significant difference between vignettes first) as between-subjects factors. The Huynh-Feldt the worried and less-worried participants for moral disap- correction was used as Mauchly’s test of sphericity was signif- proval, F(1, 484) ¼ 5.82, p ¼ .016, Z ¼ .01, such that worried icant (p < .001). Results revealed no main effects of Order p participants rated moral violations as more objectionable than F(1,483) ¼ 0.85, p ¼ .356, and no Order Group interaction, less-worried participants. Thus, political orientation was not a F(1, 483) ¼ 1.79, p ¼ .182. Therefore, order was not further confound. considered. Henderson and Schnall 7 revealed that, while moral judgments among worried partici- Contamination Disgust pants were practically identical across the three studies (Study Contamination disgust was associated with COVID-19 worry 1: M ¼ 3.42; Study 2: M ¼ 3.43; Study 3: M ¼ 3.43), the moral (r ¼ .30, p < .001), which is consistent with Studies 1 and 2. judgments for less worried participants were higher in Study 3 Following the logic of the earlier studies, we investigated the (M ¼ 3.30) than in Study 1 (M ¼ 3.18) and Study 2 (M ¼ 3.18). possible contribution of contamination disgust on moral judg- An exploratory ANOVA comparing this group in March vs. ment by performing another analysis that included the contam- May found a marginally significant effect, F(1, 272) ¼ 3.42, ination subscale of the DS-R as a covariate. This variable was p ¼ .066, d ¼ .22, such that participants in May who were again significantly related to moral condemnation, F(1, 484) ¼ relatively less worried about the disease produced harsher 77.50, p < .001, Zp ¼ .14. In contrast to the earlier two studies, moral judgments than participants less worried about the dis- however there was no longer a significant effect for corona- ease in March, perhaps as a function of extended exposure to virus worry on moral judgment after controlling for the effect this threat. Lastly, we found that the association between worry of contamination disgust, F(1, 484) ¼ .005, p ¼ .944, Zp ¼ and moral condemnation was no longer significant when con- .00, indicating that contamination concerns were largely trolling for contamination disgust, indicating that the fear of responsible for the effect descried above. pathogens was largely responsible for the effect of COVID-19 worry on moral condemnation. Contamination Disgust Over Time Because the finding that the relationship between worry and Internal Meta-Analysis moral judgments was no longer significant after controlling for Since the methods were largely the same across the three stud- contamination disgust, we considered potential reasons for this ies, we combined the data sets (N ¼ 913) to conduct a mini unexpected result. One possibility is that contamination scores meta-analysis, following the recent best-practices recommen- increased over time, as people became increasingly familiar dations of a number of researchers and statisticians (e.g., Goh with the coronavirus threat. Indeed, the effect size of the asso- et al., 2016; Lakens & Etz, 2017; McShane & Bo ¨ ckenholt, ciation between contamination disgust and COVID-19 worry 2017). Based on the guidelines proposed by Goh et al. was identical in Studies 1 and 2, but it was larger in Study 3. (2016), we used fixed effects in which the effect sizes within We therefore tested whether the participants in May reported each study and the mean effect sizes across the three studies higher contamination disgust relative to the participants tested were weighted by sample size. We first converted Cohen’s in March. A one-way ANOVA was conducted to determine if d effect sizes (Study 1: d ¼ .47, Study 2: d ¼ .42, Study 3: contamination disgust scores were different between Studies 1, d ¼ .24) into Pearson’s r for ease of analysis. All effect sizes 2, and 3. There was a main effect of contamination disgust were then Fisher’s z transformed for analyses and converted across the studies, F(2, 910) ¼ 9.63, p < .001, Z ¼ .02. Tukey back to Pearson’s r for presentation. Overall, the effect was post hoc tests revealed that contamination disgust in Study 1 (M significant, M ¼ .15, Z ¼ 4.81, p < .001, two-tailed, such that ¼ 2.63, SD ¼ .78) and Study 2 (M ¼ 2.68, SD ¼ .87) did not individuals worried about COVID-19 rated moral violations as differ, p ¼ .792 (.05, 95% CI [.13, .24]). However, the mean more objectionable than those who were less worried. A fully increase between Study 1 and Study 3 (M ¼ 2.89, SD ¼ .81) random effects test of the overall effect was also significant, as was significant (.26, 95% CI [.10, .42], p < .001), as was the indicated by a one-sample t-test of the mean effect size against mean increase between Study 2 and Study 3 (.21, 95% CI [.05, zero, M ¼ .17, t(2) ¼ 4.67, p ¼ .043, two-tailed. .37], p ¼ .004). Thus, while there was no significant increase Although all three studies showed no moderating role of between the March 17 and March 27 groups, there was a sig- foundation type, we nevertheless considered it instructive to nificant rise in contamination disgust between the March explore such potential differences, given the research interest groups and the May 6 group, suggesting that as the pandemic that the question of specificity to moral domain has received. wore on, people may have become more sensitive to contam- Purity (M ¼ .16, p < .001) showed the strongest effect, which ination, and that this concern therefore overshadowed the con- makes sense given that COVID-19 poses a direct threat to one’s tribution of coronavirus worry alone. physical health. Significant effects, however, also occurred for Harm (M ¼ .13, p < .001), Fairness (M ¼ .12, p < .001) r r Authority (M ¼ .11, p < .001) and Loyalty (M ¼ .11, r r Discussion p < .001), suggesting that the effect was relatively broad. A This preregistered study largely replicated the findings from fully random effects test of the effects against zero yielded Studies 1 and 2. What is noteworthy, however, is that the significant results for three of the five moral foundations as magnitude of the effect was somewhat smaller. One possibility indicated by one-sample t-tests against zero (all two-tailed). is the fact that the sample was more diverse in many respects, The strongest effect was for Purity, Mr ¼ .18, t(2) ¼ 5.13, given that it came from all across the U.S. In addition, it might p ¼ .036, followedbyHarm, Mr ¼ .14, t(2) ¼ 11.94, have also mattered that nearly 2 months had passed since the p ¼ .01 and Authority, Mr ¼ .12, t(2) ¼ 4.52, p ¼ .046. The outbreak of the pandemic, with many areas having issued stay- remaining two foundations, Fairness, Mr ¼ .15, t(2) ¼ 3.38, at-home orders by that point. Indeed, inspecting the means p ¼ .077, and Loyalty, Mr ¼ .13, t(2) ¼ 3.28, p ¼ .081, reached 8 Evolutionary Psychology marginal significance. Thus, the meta-analysis revealed a concern about COVID-19. Another qualification to these small-to-medium sized effect (Funder & Ozer, 2019) of worry results is the difference in the relationships between the trait- about the coronavirus on moral condemnation across different like measures of COVID-19 worry and moral judgments, and content domains. the effects of the experimental manipulation in Study 1. That is, although dispositional worry about contracting the illness was consistently related to moral condemnation, experimentally General Discussion manipulating the salience of COVID-19 had no effect on moral This research tested the role of situational concerns about an judgment, relative to a neutral condition. One possibility for infectious disease on judgments of wrongdoing. Across three why is by the time of Study 1 on March 17, news about studies we consistently found that people who were worried COVID-19 was already highly salient, and thus the experimen- about COVID-19 condemned moral wrongdoers more harshly tal manipulation did not have the intended effect. The disposi- than those who were less worried. This finding adds to emer- tional association, however, might be explained by a ging work on the role of disease threat on moral judgment. In generalized overreaction to potential harm. It is possible that Studies 1 and 2 controlling for individual differences in con- those who are prone to chronic worry about contracting an tamination disgust left the effect of coronavirus worry and infectious illness are also more sensitive to moral violations moral judgment intact. In contrast, in Study 3, we found that in disease-relevant domains as well as other moral infractions. this relationship was no longer significant after accounting for That is, fear of disease may overlap with an overgeneralized contamination disgust, indicating that fear of contamination reaction of increased sensitivity to potential harm, including was responsible for the effect. We interpret this finding to be moral wrongdoers who commit not only purity violations, but the result of a generally heightened concern about the virus at other unfavorable acts as well. Indeed, worried participants the time. Indeed, contamination disgust has been described as produced harsher judgments than less worried participants, and bearing a “striking similarity” to disease avoidance (Olatunji there was no moderating effect of moral foundation. This is et al., 2009). An intriguing possibility is, therefore, that vari- consistent with previous research, indicating that disease threat ables that are typically considered to reflect stable individual concerns are associated with conformity to moral proscriptions differences, such as disgust sensitivity, may change as a func- that are not specific to disease (e.g., Murray et al., 2011; Tybur tion of coronavirus concerns that became relatively universal et al., 2016; Wu & Chang, 2012). Lack of moderation by foun- across the world. Indeed, recent theorizing has suggested that dation type is likewise consistent with error management, such topics within the field of of psychology, and the scientific that the more costly error is to be under-vigilant about moral approaches to study them, may change in the wake of the violations that are not disease relevant than to be over-vigilant COVID-19 pandemic (Rosenfeld et al., in press). Given the solely for disease-relevant violations (Haselton et al., 2015; current findings, apart from contamination and disease con- Murray et al., 2019). Further research is needed to more care- cerns, other relevant traits such as neuroticism or conscien- fully explore these dispositional versus experimental tiousness may also have changed over the course of the differences. pandemic as a function of constantly having been engaged in Additionally, we did not test whether other variables, such disease-prevention behavior to alleviate related worries. Future as personality, might have played a role in our results. Disease research would be needed to explore this possibility. avoidance has been associated with both neuroticism and con- Our findings align with a growing body of research demon- scientiousness (Oosterhoff et al., 2018), while openness, con- strating that individual differences in the propensity to experi- scientiousness, and agreeableness have been associated with ence disgust are linked to moral considerations (Chapman & sensitivity to moral violations (Hirsh et al., 2010; Smillie Anderson, 2014; Karinen & Chapman, 2019; Liuzza et al., et al., 2020). Thus, considering the overlap between disease 2019; Murray et al., 2019; Robinson et al., 2019; Wagemans avoidance, moral judgments, and conscientiousness, this per- et al., 2018). Furthermore, the results are consistent with recent sonality trait may account for some of the variance between work showing a positive association between germ aversion worry about a highly salient communicable disease and assess- and moral condemnation across the moral foundations (Murray ments of moral wrongdoing. et al., 2019). Our findings contribute to this line of research by Our research raises the possibility that during a period of demonstrating that subjective worry about a real-world conta- widespread concern about infectious disease, people may gious disease is associated with harsher moral judgments, and, become more judgmental overall. In other words, people’s moreover, that this relationship held even after accounting for actions and intentions might be under more scrutiny, and when differences in political orientation. Thus, converging evidence ambiguous, may be interpreted uncharitably, potentially result- supports Haidt’s (2001) suggestion that morality is shaped by ing in misunderstandings, or interpersonal conflicts. Indeed, in various emotions and intuitions, of which concerns about the early days of the unfolding COVID-19 crisis, there were health and safety are prominent. media accounts of mistrust in public officials, the press, and There are limitations within these findings. Though we health organizations. The current findings suggest that we may obtained large samples with consistent results across all three see further instances of uncharitable evaluations as people are studies, we used a single item to measure “worry,” which may especially concerned for their physical health. Thus, the have reduced sensitivity in capturing participants’ level of ongoing pandemic presented an ecologically relevant way of Henderson and Schnall 9 Farrington, D. P., & Loeber, R. (2000). Some benefits of dichotomi- examining the role of disease prevalence on an issue of critical zation in psychiatric and criminological research. Criminal applied importance. Behaviour and Mental Health, 10, 100–122. Author Contributions Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G* Power R. K. H. and S. S. developed the study concept and design. R. K. H. 3: A flexible statistical power analysis program for the social, carried out the data collection and analysis, and interpreted results behavioral, and biomedical sciences. Behavior Research Methods, under the supervision of S. S. Both authors wrote the manuscript, and 39, 175–191. approved its final version. Finch, C. E. (2010). Evolution of the human lifespan and diseases of aging: Roles of infection, inflammation, and nutrition. Proceed- Open Practices Statement ings of the National Academy of Sciences, 107, 1718–1724. Studies 1 and 2 were not formally preregistered. The preregistration Finch, C. E. (2012). Evolution of the human lifespan, past, present, for Experiment 3 can be accessed at https://aspredicted.org/ and future: phases in the evolution of human life expectancy in blind.php?x¼2as8ic. Experimental materials and de-identified data relation to the inflammatory load. Proceedings of the American for all studies are available at https://osf.io/fdt6q/?view_ Philosophical Society, 156, 9–44. only¼7b4ea17bc18845d68b409b27273e82ed. Funder, D. C., & Ozer, D. J. (2019). Evaluating effect size in psycho- logical research: Sense and nonsense. Advances in Methods and Declaration of Conflicting Interests Practices in Psychological Science, 2, 156–168. The authors declared no potential conflicts of interest with respect to Goh, J. X., Hall, J. A., & Rosenthal, R. (2016). Mini meta-analysis of the research, authorship, or publication of this article. your own studies: Some arguments on why and a primer on how. Ethics Approval Statement Social and Personality Psychology Compass, 10, 535–549. Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conserva- This research follwed the ethical approval procedure of the Ethics Committee of the Department of Psychology at the University of tives rely on different sets of moral foundations. Journal of Per- Cambridge. sonality and Social Psychology, 96, 1029–1046. Haidt, J. (2001). The emotional dog and its rational tail: A social Funding intuitionist approach to moral judgment. Psychological Review, The authors disclosed receipt of the following financial support for the 108, 814–834. research, authorship, and publication of this article: The first author is Haidt, J., McCauley, C., & Rozin, P. (1994). Individual differences in supported by a Gates Cambridge Scholarship (OPP1144). Additional sensitivity to disgust: A scale sampling seven domains of disgust funding was provided by the Bennett Institute for Public Policy, where elicitors. Personality and Individual Differences, 16, 701–713. the second author is currently a Visiting Fellow. Haselton, M. G., Nettle, D., & Murray, D. R. (2015). The evolution of cognitive bias. In D. M. Buss (Ed.), The handbook of evolutionary ORCID iD psychology (2nd ed., pp. 968–987). John Wiley. Robert K. Henderson https://orcid.org/0000-0002-9869-5861 Hirsh, J. B., DeYoung, C. G., Xu, X., & Peterson, J. B. (2010). Com- Simone Schnall https://orcid.org/0000-0002-4672-7534 passionate liberals and polite conservatives: Associations of agree- ableness with political ideology and moral values. Personality and Note Social Psychology Bulletin, 36, 655–664. 1. We had not anticipated the Maine sample to be much smaller, but Karinen, A. K., & Chapman, H. A. (2019). Cognitive and personality in hindsight, it is likely that fewer people were signed up on Prolific correlates of trait disgust and their relationship to condemnation of than in Washington. As noted, the states did not differ regarding nonpurity moral transgressions. Emotion, 19, 889. coronavirus worry, and we analyzed the combined sample, so this Lakens, D., & Etz, A. J. (2017). Too true to be bad: When sets of was no longer an issue. studies with significant and nonsignificant findings are probably Supplemental Material true. Social Psychological & Personality Science, 8, 875–881. Supplemental material for this article is available online. Liuzza, M. T., Olofsson, J. K., Cancino-Montecinos, S., & Lindholm, T. (2019). Body odor disgust sensitivity predicts moral harshness References toward moral violations of purity. Frontiers in Psychology, 10,458. Chapman, H. A., & Anderson, A. K. (2014). Trait physical disgust is McShane, B. B., & Bo ¨ ckenholt, U. (2017). Single-paper meta- related to moral judgments outside of the purity domain. Emotion, analysis: Benefits for study summary, theory testing, and replic- 14, 341–348. ability. Journal of Consumer Research, 43, 1048–1063. Clifford, S., Iyengar, V., Cabeza, R., & Sinnott-Armstrong, W. (2015). Mortensen, C. R., Becker, D. V., Ackerman, J. M., Neuberg, S. L., & Moral foundations vignettes: A standardized stimulus database of Kenrick, D. T. (2010). Infection breeds reticence: The effects of scenarios based on moral foundations theory. Behavior Research disease salience on self-perceptions of personality and behavioral Methods, 47, 1178–1198. tendencies. Psychological Science, 21, 440–447. Connolly, M. A., & Heymann, D. L. (2002). Deadly comrades: War Murray, D. R., Kerry, N., & Gervais, W. M. (2019). On disease and and infectious diseases. The Lancet, 360, s23–s24. deontology: Multiple tests of the influence of disease threat on Fabrega, H. (1997). Earliest phases in the evolution of sickness and moral vigilance. Social Psychological and Personality Science, healing. Medical Anthropology Quarterly, 11, 26–55. 10, 44–52. 10 Evolutionary Psychology Murray, D. R., & Schaller, M. (2012). Threat(s) and conformity Smillie, L.D., Katic, M., & Laham, S.M. (2020). Personality and deconstructed: Perceived threat of infectious disease and its impli- moral judgment: Curious consequentialists and polite deontolo- cations for conformist attitudes and behavior. European Journal of gists. Journal of Personality. In Press. 1–16. Social Psychology, 42, 180–188. Tybur, J. M., Inbar, Y., Aarøe, L., Barclay, P., Barlow, F. K., De Barra, Murray, D. R., Trudeau, R., & Schaller, M. (2011). On the origins of M., Becker, D. V., Borovoi, L., Choi, I., Choi, J. A., Consedine, N. cultural differences in conformity: Four tests of the pathogen pre- S., Conway, A., Conway, J. R., Conway, P., Adoric, V. C., Demirci, valence hypothesis. Personality and Social Psychology Bulletin, ˇ D. E., Fernand ´ ez, A. M., Ferreira, D. C. S., Ishii, K., ... Zezel ˇ j, I. 37, 318–329. (2016). Parasite stress and pathogen avoidance relate to distinct Olatunji, B. O., Williams, N. L., Tolin, D. F., Sawchuck, C. N., Abra- dimensions of political ideology across 30 nations. Proceedings of mowitz, J. S., Lohr, J. M., et al. (2007). The disgust scale: Item the National Academy of Sciences, 113, 12408–12413. analysis, factor structure, and suggestions for refinement. Psycho- Tybur, J. M., Lieberman, D., Kurzban, R., & DiScioli, P. (2013). logical Assessment, 19, 281–297. Disgust: Evolved function and structure. Psychological Review, Oosterhoff, B., Shook, N. J., & Iyer, R. (2018). Disease avoidance and 120, 65–84. personality: A meta-analysis. Journal of Research in Personality, Van Leeuwen, F., Dukes, A., Tybur, J., & Park, J. H. (2017). Disgust 77, 47–56. sensitivity relates to moral foundations independent of political Park, J. H., & Isherwood, E. (2011). Effects of concerns about patho- ideology. Evolutionary Behavioral Sciences, 11, 92–98. gens on conservatism and anti-fat prejudice: Are they mediated by Van Leeuwen, F., Park, J. H., Koenig, B. L., & Graham, J. (2012). moral intuitions? The Journal of Social Psychology, 151, 391–394. Regional variation in pathogen prevalence predicts endorsement of Robinson, J. S., Xu, X., & Plaks, J. E. (2019). Disgust and deontology: group-focused moral concerns. Evolution and Human Behavior, Trait sensitivity to contamination promotes a preference for order, 33, 429–437. hierarchy, and rule-based moral judgment. Social and Psychologi- Wagemans, F., Brandt, M. J., & Zeelenberg, M. (2018). Disgust sen- cal and Personality Science, 10, 3–14. sitivity is primarily associated with purity-based moral judgments. Rozin, P., Haidt, J., & McCauley, C. R. (2008). Disgust. In M. Lewis Emotion, 18, 277–289. & J. M. Haviland (Eds.), Handbook of emotions (3rd ed., World Health Organization. (2015). World health in 2015: From pp. 757–776). Guilford Press. MDGs, millennium development goals to SDGs, sustainable devel- Schnall, S. (2017). Disgust as embodied loss aversion. European opment goals. https://apps.who.int/iris/handle/10665/200009 Review of Social Psychology, 28, 50–94. Wu, B. P., & Chang, L. (2012). The social impact of pathogen threat: Schnall, S., Haidt, J., Clore, G. L., & Jordan, A. H. (2008). Disgust as How disease salience influences conformity. Personality and Indi- embodied moral judgment. Personality and Social Psychology vidual Differences, 53, 50–54. Bulletin, 34, 1096–1109. Zmigrod, L., Ebert, T., Go ¨ tz, F. M., & Rentfrow, J. (2021). The psy- Shaw-Taylor, L. (2020). An introduction to the history of infectious chological and socio-political consequences of infectious diseases. diseases, epidemics and the early phases of the long-run decline in In Press. mortality. The Economic History Review, 73, E1–E19.
Evolutionary Psychology – SAGE
Published: Jun 10, 2021
Keywords: morality; disgust; pathogen avoidance; behavioral immune system; moral judgment; emotion; harm; COVID-19; coronavirus; moral foundations theory
Access the full text.
Sign up today, get DeepDyve free for 14 days.