Make No Apologies: Fear of Negative Evaluation, Depressive Symptoms, and the Mediating Role of Accounting for COVID-Safe Behavior Amongst People at High-Risk for Severe Illness

ABSTRACT

With the goal of understanding unique and important threats to the mental health of people who are especially vulnerable to severe illness as a result of COVID-19, this study investigated associations between such individuals’ fear of negative evaluation, tendency to “account for” practicing COVID-safe behaviors, and depressive symptoms. Grounded in perspectives on self-presentation, normative influence, and cognitive dissonance, we hypothesized that fear of negative evaluation would relate positively to accounting for COVID-safe behaviors, which, in turn, would associate positively with increased depressive symptoms. The results showed that increased fear of negative evaluation predicted an increased use of apologies and excuses, which in turn were positively related to depressive symptoms. Justifications for COVID-safe behaviors were not significantly associated with either fear of evaluation or depressive symptoms. The practical and theoretical implications are discussed.

Considering COVID-19’s profound effects on every part of society, the etymological origins of the word “pandemic”—derived from the Greek pan, meaning “all,” and demos, meaning “people” (Honigsbaum, 2009)—seem most appropriate. And yet, in the midst of such widespread effects, different groups of people experience the pandemic in unique ways (e.g., see Zhou et al., 2022). One group of people whose experiences are crucial to understand are those who are high risk for complications stemming from COVID-19. Indeed, people with certain underlying health conditions are not only at elevated risk for severe illness, hospitalization, and death from the virus that causes COVID-19 (e.g., Guo et al., 2020; Liang et al., 2020), they suffer from increased mental health issues as a result (Al-Rahimi et al., 2021). According to Islam et al. (2021) such mental health symptoms may not only be due to increased fear of contracting the disease, but also may result from an increased adherence to COVID-19 preventative measures, including social distancing, which are associated with increased social isolation and lack of support.

As the Centers for Disease Control and Prevention (CDC) no longer advises indoor mask-wearing for most people, as various mandates have lifted, and as adherence to preventative measures becomes less common, the situation facing people with certain underlying health conditions has become uniquely stressful in many respects. Not only do such individuals still worry about infection, empirical evidence suggests that self-presentational concerns further threaten their mental well-being. Specifically, as preventative measures such as mask-wearing become less prevalent, high-risk individuals may feel pressure to fit in. In one study (Wilson et al., 2020), for example, young adults reported receiving social pressure from others to not wear masks and feeling weird when wearing masks. Similarly, Chinese students in the UK reported feeling stressed and embarrassed when wearing facemasks in public, due, in part, to social norms against such behavior (Lai et al., 2021). To further complicate matters, recent research suggests that mask mandates have been ineffective at reducing the spread of COVID-19 at the population level (Jefferson et al., 2023).¹ To the extent that such findings risk being misinterpreted as meaning that masks do not protect individuals–when, in fact, the correct masks worn properly do (e.g., see Ueki et al., 2020)–people who are at high risk could face even more social pressure when attempting to protect themselves.

Considering this, previous theory and research suggests that people with underlying conditions may feel compelled to account to others for their COVID-safe behaviors. Specifically, theories of impression management (Goffman, 1959, 1971) and politeness (Brown & Levinson, 1987) argue that people seek to maintain desirable images. As such, to avoid losing face after engaging in questionable behaviors (e.g., anti-normative actions, failures, transgressions), people often “account for” such behaviors through the use of explanations, including excuses, justifications, and apologies. Guided by these theoretical tenets, the present study explores how accounting techniques for COVID-safe behaviors mediate the relationship between high-risk individuals’ fear of negative evaluations and depressive symptoms. In doing so, our model highlights the social factors that put high risk individual’s mental health in danger in a time where their physical health is already in jeopardy.

Review of literature

Normative influence in the context of COVID-19

Nearly a century of research on social influence indicates that people often change their attitudes, beliefs, and behaviors to match what others do (descriptive norms) and approve of doing (injunctive norms) (Asch, 1956; Cialdini & Goldstein, 2004; Sherif, 1936). According to such research, the tendency to conform is prompted by two underlying motivations: the desire to be right and the desire to be liked (Campbell & Fairey, 1989; Cialdini & Goldstein, 2004; Deutsch & Gerard, 1955). In the first case, conformity results from informational influence. Specifically, other people are perceived as a source of information about what is most accurate or appropriate. In the second case, conformity results from normative influence, which is to say that people follow others in order to fit in, gain acceptance, and avoid disapproval. In Solomon Asch’s (1956) classic experiments, for example, over one-third of participants provided obviously incorrect answers to a line-judgment task in order to fit in with a group of confederates. Follow up interviews indicated that self-doubt and normative pressure played significant roles in conformists’ decision to go along with the group. Relatedly, in one study (Janes & Olson, 2000), participants who viewed others being ridiculed were subsequently more likely to fear failure and conform to social norms.

Although early literature (e.g., Goffman, 1963) pointed to negative norm violations as a source of ridicule, more contemporary work has explored the experiences of people who undergo stigmatization as a result of diverging from group norms by engaging in healthy behaviors (Romo, 2012, 2018; Romo & Donovan Kicken, 2012). For example, former problem drinkers who abstain from alcohol (Romo et al., 2016) and individuals who have lost weight (Romo, 2018) often feel stigmatized and rejected by people in their social networks who may resent or feel threatened by such healthy behavior. Likewise, vegetarians face unique pressures when attempting to fit in with mainstream society while pursuing behaviors that others view as unconventional (Romo & Donovan Kicken, 2012).

Along these lines, we suggest that people practicing COVID-safe behaviors may experience similar challenges. Indeed, according to Packer et al. (2021), people’s tendency toward conformity is particularly relevant in the context of the COVID-19 pandemic. Not only is there higher uncertainty and need for affiliation – which increase informational and normative influence, respectively – people are motivated to embody the identity of their own rather than other groups (Packer et al., 2021). As a result, to the extent that others practice COVID-safe behaviors, normative factors might prompt individuals to follow suit (e.g., see H. J. Kim & Han, 2022; Kittel et al., 2021). Alternatively, however, when others engage in less-safe behaviors, learning about normative patterns can backfire, pressuring individuals to conform (Packer et al., 2021).

Previous literature points to a number of factors that might prompt such normative pressure. By way of example, despite the demonstrated effectiveness of many safety measures, putting them into practice has become one of the most politicized and contentious issues in the United States (Kwon, 2022). Republicans, for instance, are less likely than Democrats to practice COVID-safe behaviors, and conservative messaging has often framed preventative guidelines as a threat to personal freedom (Pascual-Ferrá, 2021; Young et al., 2022). In reaction to various mandates during the pandemic, some groups took to the streets in protest, blocked access to medical facilities, and harassed health care workers. One analysis indicated that such groups are characterized by high levels of cohesion and pressure to conform (Forsyth, 2020). Considering this, it is not surprising that some people have chosen to get vaccinated in secret, fearing backlash and disapproval from family and friends who shun COVID-safe practices (Mastroianni, 2021).

Of course, not all preventative measures are hidden so simply. As Bir and Widmar (2021) noted, mask wearing in public is visually observable and easily responded to through shaming and ostracizing. Even in the absence of such overt antagonism, people might experience normative pressure while practicing COVID-safe behavior. For example, many people in Western cultures worry that they will look strange or be judged as strange when practicing COVID-safe behaviors (Friedman, 2020; Rieger, 2020). Some believe that wearing masks looks shameful, uncool, and weak (Capraro & Barcelo, 2020).

Additional research indicates that the mere presence of people not wearing masks can increase feelings of normative pressure. Woodcock and Schultz (2021), for instance, reported that mask-wearing, or lack thereof, appears to be influenced by the behavior of proximal others (Woodcock & Schultz, 2021), while Carbon (2021) found that people imagined feeling more “strange” about wearing a mask in social settings when the frequency of non-masked others in the settings increased (Carbon, 2021). In short, previous literature suggests that, as fewer individuals are practicing safe behaviors in the United States (Reimann, 2022), people who want to continue such practices may feel awkward or alienated when doing so.

Self-presentation and accounting for COVID-safe behaviors

Recently, a qualitative study (Ma & Zhan, 2022) of Chinese university students in the U.S. found that, despite extreme anxiety over getting sick, several students, including those who had endured harassment over mask-wearing from strangers on the street, removed their masks in the classroom out of concern for how they would be perceived by classmates. Such experiences are consistent with theoretical work on impression management (Goffman, 1959, 1971), strategic self-presentation (Jones & Pittman, 1982; Leary, 1995), and politeness (Brown & Levinson, 1987). Specifically, such theories argue that people are motivated to create and maintain desirable public images (i.e., “positive face”). As such, they selectively present (and/or omit) information about themselves to influence perceptions of themselves by others. Previous literature notes that self-presentational motives affect a wide variety of behaviors, including conformity (Cialdini & Goldstein, 2004). As Carron et al. (2004) noted, by yielding to group norms, people avoid the embarrassment of seeming different and are more likely to be accepted by the group.

But what if people desire acceptance, yet engage in behavior that violates social norms? Although the lion’s share of impression management research has focused on “assertive” actions, which people use proactively to achieve their self-presentational goals, theorists have also identified “defensive” actions, which people use to protect desired images (Schlenker, 1980; Tedeschi & Riess, 1981). Such defensive actions include offering explanations that “account for” norm-violating behaviors (e.g., Benoit & Strathman, 2004; Bies et al., 1988; Goffman, 1971; Greenberg, 1990; Holtgraves, 1989; McLaughlin et al., 1983; Schonbach, 1980; Scott & Lyman, 1968). Although there is no generally accepted typology of accounts, most include apologies, excuses, and justifications as prominent types of defensive explanations. First, when communicating an apology, a person acknowledges being responsible for norm-violating behavior and expresses penitence. Second, in offering an excuse, a person shifts responsibility for a questionable behavior from themselves to some external cause. Third, in presenting a justification, a person admits being responsible for a behavior, but claims that it was warranted.

Considering such research alongside evidence that, in some contexts, practicing COVID-safe behavior is perceived as anti-normative (see above), we suspected that some high-risk people will offer explanations in order to “account for” such behaviors. Previous research in other contexts supports this proposition. Specifically, research has found that in order to fit in with people in their social networks, stigmatized nondrinkers (Romo, 2012; Romo et al., 2016), vegetarians (Romo & Donovan Kicken, 2012), and people who have lost weight (Romo, 2018) feel pressure to smooth over social interactions, often through the use of explanations (e.g., excuses). With that in mind, we examine how people who are at increased risk for severe illness might “account for” their COVID-safe behavior (e.g., social distancing, mask wearing and so forth).

The present study

Although self-presentation is ubiquitous, not all people are equally motivated to seek approval. For example, people who fear negative evaluation (Watson & Friend, 1969) worry about others’ disapproval, and as such, work harder and focus more on fitting in and being liked than people who worry less (Griskevicius et al., 2006; Leary & Allen, 2010; Nezlek & Leary, 2002). With regard to COVID-19, Carlton et al. (2022) suggested that fear of negative evaluation might lead people to ignore safety guidelines in locations where such guidelines are disapproved of. Alternatively, in circumstances where such people choose to remain safe, thereby diverging from group norms, what might they do to avoid disapproval? As noted above, theories of impression management (Goffman, 1959, 1971) and politeness (Brown & Levinson, 1987) argue that people explain their questionable behavior (e.g., anti-normative actions, failures, transgressions) by offering accounts, which include excuses, justifications, and apologies. Considering this, we tested the following hypothesis:

H1:

Fear of negative evaluation will be positively associated with acts of accounting for COVID-safe behavior.

There is reason to expect that accounting for COVID-safe behaviors will be negatively associated with account-givers’ psychological well-being. Specifically, previous literature indicates that people with conditions that make them vulnerable to COVID-19—including cancer (Huang et al., 2021), obesity (Hill et al., 2021), chronic respiratory disease (Brighton et al., 2022), and diabetes (Schabert et al., 2013)—often feel stigmatized as a result of their condition. To avoid feelings of shame and rejection, such stigmatized people often hide their condition from others (see Ma & Zhan, 2022). Whiddett et al. (2006), for example, noted that people are increasingly reluctant to share personal health information with anyone other than their doctors. Even then, people may withhold information from health professionals out of a concern for privacy (Fox, 2020). In one study, a majority of interviewees expressed a strong desire for privacy as underlined by comments such as “I place utmost value on the privacy of my health information […] it can have an impact on many aspects of your life, you don’t want it getting into the wrong hands” (Fox, 2020, p. 1021).

These findings are consistent with extant theory and research. First, communication privacy management (CPM) theory (Petronio, 2002, 2007, 2013), argues that decisions about whether to reveal or conceal information to others are characterized by tension. Indeed, there are risks associated with sharing private information. As such, people rely on a system of boundary rules to decide when, how, and with whom to share such information. That said, external situational factors can trigger a shift in criteria (known as catalyst criteria), leading to a new set of privacy rules. We suggest that the COVID-19 pandemic represents one such factor. That is, to the extent that others refrain from engaging in behaviors that protect vulnerable people (e.g., wearing a mask), the latter may feel compelled, contrary to their concerns for privacy, to disclose their personal health information in order to account for their COVID-safe behaviors. In a related study, for example, when students with learning disabilities wanted to be viewed positively, they disclosed their disability to classmates as a way to explain their nonconforming behavior (e.g., walking around the classroom to maintain focus; Johnsen et al., 2017).

Second, a key assumption of cognitive dissonance theory (Festinger, 1957) is that incongruities between beliefs and/or freely-chosen behaviors cause people to experience psychological discomfort, which, in turn, is associated with negative affect, including depression (e.g., see Elliot & Devine, 1994; Harmon-Jones, 2000; Hopwood, 2022). In the case of high-risk individuals accounting for their COVID-safe behaviors, such incongruities seem likely. Indeed, if practicing COVID-safety is appropriate and correct (otherwise, why wear masks at the risk of social disapproval?), then the act of defending such practices seems contradictory. Moreover, to the extent that making difficult decisions induces dissonance (see Harmon-Jones & Mills, 2019), the choice between saying nothing (in order to protect one’s private health information) or offering accounts (in order to defend one’s correct behavior) might also present risks to people’s mental well-being. With this in mind, we tested the following hypothesis:

H2:

Accounting for COVID-safe behavior will be positively associated with feelings of depression.

H3:

There will be indirect effects from fear of negative evaluation to depressive symptoms through accounting for COVID-safe behaviors.

Method

Participants and procedure

Data collection occurred from July – September of 2022 using Amazon’s Mechanical Turk (MTurk). Participants were eligible if they were 18 years or older, living in the U.S., and identified themselves as being at increased risk of becoming severely ill as a result of COVID-19. To be eligible, participants had to check at least one of the listed medical conditions that, according to the CDC, make someone high risk for complications stemming from COVID-19. Participants were allowed to check as many as applied to their current medical condition. Participants also must have completed 1,000 human-intelligence tasks with a minimum 98% success rate as an MTurk worker. Participants completed online surveys assessing their depressive symptoms, fear of negative evaluation from others, and accounting for COVID-safe behaviors as well as variables not included in this study such as social support. At the end of the survey, participants received a code to record in their MTurk portal to receive $1.50. Each measure below was taken at one time point. The IRB approved all procedures for this study.

This sample – ranging in age from 21–74 years (M = 40.80, SD = 11.10)—consisted of 126 participants reporting male (46.5%), 132 reporting female (48.7%), and 3 people reporting gender nonconforming 1.1%) with 10 people not reporting. The reported race/ethnicity of the sample were as follows: 14 Black/African American, 16 Hispanic/Latinx, 193 White/Caucasian, 30 Asian/Pacific Islander, 4 Native American/Alaskan Native, and 3 Other with 11 people not reporting. Table 1 shows a frequency table of participant’s listed medical condition that made them higher risk individuals.

Table 1. Reports of medical conditions from participants.

Condition	Frequency	Percentage of the Sample
Cancer	36	13.3%
Chronic kidney diseases	30	11.1%
Chronic liver diseases	35	12.9%
Chronic lung diseases	41	15.1%
Cystic fibrosis	14	5.2%
Dementia or other neurological conditions	15	5.5%
Diabetes (types 1 and 2)	72	26.6%
Disabilities	36	13.3%
Heart conditions	45	16.6%
HIV infection	14	5.2%
Immunocompromised condition or weakened immune system	36	13.3%
Mental health conditions	58	21.4%
Overweight and obesity	99	36.5%
Physical inactivity	53	19.6%
Pregnancy	16	5.9%
Sickle cell disease or thalassemia	14	5.2%
Smoking, current or former	71	26.2%
Solid organ or blood stem transplant	7	2.6%
Stroke or cerebrovascular disease	12	4.4%
Substance use disorders	20	7.4%
Tuberculosis	13	4.8%
Other (specify if so)	16	5.9%

The percentages total to over 100% because participants were allowed to select as many as applied to their medical condition. People who selected “Other” typed in conditions such as Asthma and Crohn’s Disease. For more information, please contact the authors.

Measures

Fear of negative evaluation

The Brief Fear of Negative Evaluation Scale (Leary, 1983) includes 13-items (e.g., “I am usually worried about what kind of impression I make” and “I am afraid that people will find fault with me) measured on a 5 – point Likert scale (1 = not at all characteristic of me to 5 = extremely characteristic of me). Higher scores indicate higher fear of negative evaluation (α = .90; M = 2.96; SD = .87).

Accounting for COVID-safe behaviors

The Self-Presentation Tactics Scale (Lee et al., 1999) measures the use 12 self-presentation tactics, including apologies, excuses, and justifications. We adapted appropriate items from each of these three subscales in order to focus on COVID-specific accounts. The resulting items were measured on a 5– point Likert scale (1 = Strongly disagree to 5 = Strongly agree) with higher scores indicating higher agreement that a tactic was used. The apology subscale was measured with three items (e.g., “I apologized to others when practicing COVID-safe behaviors) (α = .92; M = 2.57; SD = 1.30), excuses were measured with three items (e.g., “While practicing COVID-safe behavior, I used my underlying illness as an excuse”) (α = .86; M = 2.72; SD = 1.21), and justifications were measured using 4 items (e.g., “While others might have viewed my COVID-safe behaviors as negative, I offered my vulnerable health condition as an explanation so that they would understand that my behavior is justified.”) (α = .93; M = 3.02; SD = 1.21).

Depressive symptoms

Depressive symptoms were measured using Santor and Coyne’s (1997) Revised Center of Epidemiologic Studies Depression Scale (CES-D). The scale consists of 8 items (e.g., “I felt depressed.” “I felt sad.”) measuring the frequency of depressive symptoms over the past week and rated on a 4-point Likert scale (1 = less than 1 day to 4 = 5–7 days). Higher scores show higher levels of depressive symptoms (M = 2.55, SD = 1.11, α = .95).

Desire for medical privacy

We measured one’s desire for medical privacy by adapting three items from the Informational Privacy subscale of Trepte and Masur’s (2017) Need for Privacy Questionnaire. This adapted subscale consisted of 3 items (e.g., “I prefer that not much is known by others about my personal health.”) and it is measured on a 5-point Likert scale (1 = Strongly disagree to 5 = Strongly agree). Higher scores indicate more desire for medical privacy (α = .81; M = 3.92; SD = .90).

Results

Table 2 shows the correlations amongst the variables in the study. Of note, fearing negative evaluations from others was positively associated with excuse making and apologies when accounting for COVID-safe behaviors as well as depressive symptoms. Moreover, each accounting behavior was positively linked to depressive symptoms.

Table 2. Correlations among study variables.

	1	2	3	4	5
1. Fear of Negative Evaluation
2. Justification	.07
3. Excuses	.20**	.76**
4. Apology	.23**	.66**	.81**
5. Depressive Symptoms	.52**	.36**	.56**	.57**
6. Need for Medical Privacy	−.08	−.05	−.09	−.16*

*p < .05; **p < .01.

The hypothesized model was tested using Hayes’ (2013) PROCESS 3.0 Macro for SPSS. Model 4 of the macro allows for a simultaneous test of parallel mediators from the exogenous variable to the endogenous variable. The data presented below are cross-sectional and do not reflect causal relationships amongst the variables. The output produced coefficients from fear of negative evaluation (X) to depressive symptoms (Y; c’ path), while controlling for the mediating variables. It also produced paths from fear of negative evaluation (X) to all three accounting behaviors: justifications (M₁; a₁ path), excuses (M₂; a₂ path), and apologies (M₃; a₃ path). Moreover, the macro tests the paths from each account behavior: justifications (M₁;b₁ path), excuses (M₂; b₂ path), and apologies (M₃; b₃ path) to depressive symptoms (Y), and finally, all three indirect effects from fear of negative evaluations (X) depressive symptoms (Y) through the three accounting behaviors (ab paths). The mediating variables controlled for each other in the analysis. We also added self-reported desire for medical privacy as a control variable in the model. Given that each accounting technique deals with communicating about one’s medical condition, one’s desire to keep that information private could be a rather large confounding variable. Thus, we account for the variance of one’s need for medical privacy in the model. As such, the results reflect the links between fear of negative evaluations, accounts, and mental health above and beyond one’s need for medical privacy. The macro generated 95% bias corrected and adjusted confidence intervals (CI). CIs that exclude zero indicate significant indirect effects.

The unstandardized regression coefficients for the hypothesized model are depicted in Figure 1. H1 predicted a significant positive association between fear of negative evaluations and all three accounting behaviors. As shown, the results revealed no significant effect from fear of negative evaluations to justification (B = .07, SE = .10, t = −.69, p > .05), however, there were significant positive association for excuses (B = .26, SE = .10, t = 2.75, p < .01) and apologies (B = .35, SE = .10, t = 3.50, p < .001). Thus, H1 was partially supported.

Figure 1. The hypothesized model.

Note. H3 tested the indirect paths from fear of negative evaluations to depressive symptoms through each mediating variable. The model also controlled for desire for medical privacy.

Next, H2 predicted positive relationships between each accounting behavior and depressive symptoms. The results showed no significant association from justification to depressive symptoms (B = −.09, SE = .07, t = −1.33, p > .05). There were significant positive associations between excuses (B = .32, SE = .09, t = 3.56, p < .001) and apologies (B = .24, SE = .07, t = 3.34, p < .01) and depressive symptoms. Thus, H2 was also partially supported.

H3 tested the indirect associations between fear of negative evaluations and depressive symptoms through each accounting behavior. Below we report the completely standardized indirect effect of each mediation path as well as the SE and confidence intervals. There was no significant indirect effect through justification (ß = −.005, SE = 0.01, 95% CI: −.04, .01). There was a significant positive indirect effect from fear of negative evaluation to depressive symptoms through excuses (ß = 0.07, SE = .03, 95% CI: .01, .14) and apologies (ß = .07, SE = .03, 95% CI: .02, .14). H3 was also partially supported.

Discussion

In an effort to understand unique and important threats to the mental health of people who are especially vulnerable to severe illness as a result of COVID-19, this study investigated associations between individual differences, the tendency to “account for” practicing COVID-safe behaviors, and depressive symptoms. Specifically, grounded in theories of self-presentation and politeness, we posited that fear of negative evaluation would relate positively to offering accounts, which, in turn, would relate positively with increased suffering from depressive symptoms. The results partially supported these hypotheses.

First, as suspected, our results indicated that COVID-vulnerable people who feared negative evaluations were more likely to apologize and make excuses in order to account to others for practicing protective behaviors. This finding is consistent with previous theory and research suggesting that people who are especially motivated to avoid social disapproval differ with regard to how they attempt to influence others’ impressions (see Leary & Allen, 2010). Contrary to our expectations, however, people’s fear of negative evaluation was not associated significantly with accounting for COVID-safe behavior through the use of justifications. Previous literature derived from theories of politeness and causal attribution invite one possible explanation for these results. Specifically, McLaughlin, Cody, and their colleagues (Cody & McLaughlin, 1990; McLaughlin et al., 1983) theorized a mitigation-aggravation continuum, along which several types of accounts could be ordered with respect to how effectively they function to neutralize receivers’ anger and negative evaluations of a source. At one end of the continuum, apologies and excuses are considered the most mitigating types of account. At the other end, justifications (and denials) are considered the most aggravating because, in claiming that a questionable behavior is warranted, justification threaten the face needs of people who are receiving the account (Gonzales, 1992). With that in mind, to the extent that justifications, relative to excuses and apologies, prompt anger and disapproval, it stands to reason that people who fear negative evaluation would avoid using them.

A related explanation might also explain why, contrary to our second hypothesis, justifications were not positively associated with depressive symptoms, although, consistent with the hypothesis, apologies and excuses were. Specifically, in terms of threatening (or protecting) a person’s face needs, mitigating and aggravating accounts are often thought to have opposing effects on sources and receivers (Gonzales, 1992). That is, because apologies accept responsibility and acknowledge blameworthiness, they, unlike justifications, aim to protect the face needs of others at the expense of the person providing the account. In contrast, because justifications deny that a behavior was wrong or harmful, they, unlike apologies and excuses, aim to protect the face needs of the person providing the account at the expense of others (Gonzales, 1992; Schonbach, 1990). As a result, people who offer justifications may be less prone to depressive symptoms than those who apologize or make excuses for their COVID-safe behavior. In fact, to the extent that justifications affirm or validate behaviors in question, future research should explore their effects on receivers’ perceptions of account-givers and COVID-safe behaviors.

In addition to such considerations, the results of this study point to important research implications. Specifically, to our knowledge, this is the first study to document that some people who are at high-risk for severe illness, hospitalization, and death from the virus that causes COVID-19 offer accounts to apologize or make excuses for practicing the very behaviors that protect them from such risks. As noted above, we suggest that concerns over self-presentation, fitting in, and cognitive dissonance are likely explanations for our results. An alternative explanation – i.e., that offering such accounts threatens one’s desire for medical privacy – was controlled for in our analysis. Thus, we were able to show that the way in which people account for their behavior is related to their mental health above and beyond the confounding explanation that an account impedes a person’s ability to keep their medical information private. This highlights the importance of understanding how cognitive dissonance and self-presentation concerns are important factors for mental well-being amongst high risk individuals.

Moreover, our results highlight important theoretical considerations for research on accounts. Specifically, although previous qualitative studies (see above) indicate that people account for non-normative healthy behavior in order to fit into social groups, our results extend such research by examining the effects of such accounting behavior. Moreover, although studies in other contexts have examined the mental-health consequences of providing accounts, almost all have focused on the effects of excuses, generally pointing to benefits for excuse makers, including higher self-esteem and lower levels of depression (for a review, see Snyder & Higgins, 1988). The underlying explanation for such findings is that excuses shift causal attributions for negative outcomes from internal to external sources and, in so doing, protect a person’s positive sense of self (Snyder & Higgins, 1988). Our study, in contrast, challenges such prevailing views, suggesting that the advantages of offering accounts become suspect in contexts where communicating such accounts may contradict the beliefs and better judgment of the account-maker.

In addition to its academic contributions, this study holds important applied implications for people who are vulnerable to COVID-19, and for those who are not. First, previous research in organizational contexts indicates that, although feeling accountable to others can have positive effects, it can also lead to conformity, even when doing so results in negative personal outcomes, including poor decision-making (see Hall et al., 2006; Quinn & Schlenker, 2002). With that in mind, it is not surprising that a sense of increased autonomy has been found to neutralize the need to account for one’s actions (Hall et al., 2006). Considering this alongside associations between autonomy and social support (e.g., B. Kim et al., 2019), efforts to promote autonomy provide one possible avenue for helping high-risk individuals cope with social pressures in the context of COVID-19. Previous research (Quinn & Schlenker, 2002), for instance, demonstrated that when people expected to account for their decisions, those primed with the goal of being accurate (i.e., informative influence) displayed more independence and made better decisions than those primed with goal of fitting in (i.e., normative influence). Such research underlines the continued importance of reinforcing high-risk individuals’ beliefs in the effectiveness of COVID-safe behaviors while highlighting the potential pitfalls of seeking approval from groups of people who lack knowledge or concern.

That said, it would be shortsighted of us to suggest that vulnerable people simply ignore their motivation to account to others. Indeed, as we have argued, humans have social needs – particularly in the context of COVID-19 (Lai et al., 2021). Thus, the notion that they should ignore these inclinations may cause more harm than good. Considering this, a straightforward inference from our results suggests that high-risk individuals might consider constructing “social scripts” to explain (if they want to) why they are engaging in particular COVID-safe behaviors (e.g., social distancing, mask wearing, attending a meeting via video chat). In other words, although high-risk individuals should feel no pressure to account for their evidence-based COVID-safe behaviors, should they choose to do so, results of this study indicate that they should make no apologies or excuses, which were associated with depressive symptoms. Instead, they might focus on justifications, which were not linked to negative outcomes in this study (but also see below). Constructing justificatory social scripts before a potentially stressful interaction could go a long way in mitigating the negative effects of apologizing or making excuses for COVID-safe behaviors. High-risk people may benefit from planning out a reason for their behavior in a way that not only makes them feel more confident, but also reinforces why they are engaging in a particular behavior. Moreover, social scripts might lessen the distress associated with accounting for one’s behavior, as they allow people to avoid coming up with a compelling reason in the moment.

Although we believe the implications expressed above would help high risk individuals manage social situations in ways that would potentially stave off mental health risks, they do not address the responsibility of the receivers of accounts for COVID-safe behaviors. Clearly, individuals who are high risk for COVID-19 might account for their logical and healthy choices because they fear others will perceive them negatively. Given that vulnerable people for COVID-19 already deal with the stress of immunocompromised health issues, non-vulnerable people could work to dismantle the notion that COVID-safe behaviors are in violation of any social norms. Put simply, non-vulnerable individuals could work to make vulnerable people feel comfortable, accepted, and welcomed regardless of their COVID-safe choices. This strategy could be useful considering previous research, which indicates that affirming one’s values and emphasizing the protection of others are effective approaches for promoting COVID-safe behaviors (Gillman et al., 2022). Moreover, we suggest that rather than questioning or casting doubt on someone’s COVID-safe behaviors (e.g., “Why are you wearing a mask?”), a better social response could be to participate in the behavior (e.g., also wear masks), or at the very least, communicate respect and acceptance for someone’s choice to engage in a safe and healthy practice. Bouman and Steg (2021) found that caring about “distant others” is an important factor in predicting COVID-safety. Our findings suggest that increasing such public concern about general vulnerable populations might also help reduce social stress for high risk groups. Overall, these recommendation are consistent with previous research pointing to the importance of social support when resisting normative influence. In Asch’s (1956) classic experiments, for example, if one confederate expressed disagreement with the majority’s incorrect choice, participants’ conformity decreased by as much as 80%.

Despite the contributions of this study, it is not without limitations. For example, the cross-sectional nature of our data do not allow for causal inferences. It is possible, for example, that depressive symptoms could influence one’s fear of negative evaluations from others. Moreover, our sample predominantly contained people who identified as White/Caucasian. Given the robust history of discrimination amongst marginalized groups in the context of COVID-19 a different sample of high-risk individuals may have yielded different results. Our model also does not account for some variables that could be quite relevant to one’s perception of COVID-safe behaviors such as political affiliation (see Kwon, 2022). Political affiliation is indeed linked to COVID-related behaviors and thus, future research should consider how political identities interact with COVID related outcomes. Last, our data reflect the (dis)advantages of using Amazon’s Mechanical Turk. As Aguinis et al. (2021) note, Mturk is a flexible and convenient tool for collecting a large sample. Yet, our study should be read in light of the limitations of Mturk such as the potential for participant inattentiveness, misrepresentation, and the prevalence of bots. In addition to taking such issues into account, future research might consider audience reactions to various accounts for COVID-safe behaviors. For instance, to the extent previous research indicates that justifications tend to be more aggravating to audiences than apologies and excuses (McLaughlin et al., 1983), might their use backfire, further threatening the mental well-being of high-risk individuals? Moreover, future research could examine if people use justification as a subtle way of persuading others to engage in the COVID-safe behaviors they are accounting for. Future research should also consider an array of other variables, including the effect of cultural differences. Specifically, because conformity is both socially prescribed and normatively welcomed in some cultures more than others (Cialdini & Goldstein, 2004), and because, historically, some cultures devalue mask wearing more than others (Kwon, 2022), might members of such cultures respond differently to accounts surrounding COVID-19? Overall, by focusing on people who are especially vulnerable to COVID-19, this research contributes to understanding unique health risks facing certain populations. In addition to underlining important theoretical considerations, we hope this study provides practical guidance to people confronting complicated, and often competing, concerns for self-presentation and self-protection.

Ethical approval

This research was approved by the IRB of Utah State University, Protocol 12871.

Acknowledgments

We thank Debora Seiter for her insights.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by a grant from Utah State University’s Faculty-Student Summer Mentorship Grant.

Notes

1. Although wearing masks protects individuals, mask mandates may have been ineffective at the population level, in part, because people disregarded mandates or did not wear masks properly (see Leonhardt, 2022).