We're all in this together: Focus on community attenuates effects of pandemic-related financial hardship on reactance to COVID-19 public health regulations.

There has been resistance to COVID-19 public health restrictions partly due to changes and reductions in work, resulting in financial stress. Psychological reactance theory posits that such restrictions to personal freedoms result in anger, defiance, and motivation to restore freedom. In an online study (N = 301), we manipulated the target of COVID-19 restrictions as impacting self or community. We hypothesized that (a) greater pandemic-related financial stress would predict greater reactance, (b) the self-focused restriction condition would elicit greater reactance than the community-focused restriction condition, (c) reactance would be greatest for financially-stressed individuals in the self-focused condition, and (d) greater reactance would predict lower adherence to social distancing guidelines. Independent of political orientation and sense of community, greater financial stress predicted greater reactance only in the self-focused condition; the community-focused condition attenuated this association. Additionally, greater reactance was associated with lower social distancing behavior. These findings suggest that economic hardship exacerbates negative responses to continued personal freedom loss. Community-focused COVID-19 health messaging may be better received during continued pandemic conditions.

Introduction

Government-mandated or encouraged public health regulations to decrease the community transmission of the COVID-19 virus have elicited diverse reactions from the populace of the U.S. The effectiveness of such regulations for decreasing transmission is contingent upon the acceptance and cooperation of the population. Whereas many have abided, others have refused these regulations or guidelines as illegitimate and inappropriate government overreach infringing upon their rights [1, 2]. Why, and what can be done to reduce reactance against public health measures in the face of a global pandemic? Psychological reactance theory (PRT) posits that restriction or loss of behavioral freedoms results in anger, negative cognitions, and motivation to restore freedoms [3-5]. We propose that adults who have experienced greater economic hardship due to COVID-19 public health regulations will evince greater reactance, but that focusing their perspective on the regulations’ impact on their community–rather than personal impact–will attenuate that reactance.

Reactance in the pandemic context

PRT proposes that maintaining agency and autonomy, known as behavioral freedom, is a universal human motivation [6]. Restricting or eliminating behavioral freedom generates reactance [3, 4], both when freedoms are explicitly restricted (e.g., state-enforced curfew) as well as when restrictions are implied (e.g., messages to limit one’s own behavior; [3, 7]. Public health campaigns to reduce unhealthy behavior and to promote health behavior have been found to generate reactance [8-11]. This may be due to messaging being seen as threatening to personal freedom for health behaviors, as the magnitude of reactance experienced is associated with the freedom’s perceived importance and extent of threat [5]. As such, PRT presents a potential explanation for resistant or defiant responses in the COVID-19 pandemic context. Public health regulations that were experienced as restricting or removing highly-valued personal freedoms would have been experienced as threatening, and therefore likely to engender greater reactance. For example, regulations such as social distancing, limited capacity in stores and workplaces, and designations of (non-)essential work status may directly impact individuals’ abilities to meet routine life demands such as paying rent or mortgage, buying food and essential supplies, and supporting other financial needs [12]. The effects may be particularly damaging when the restrictions to behavioral freedom have the potential to be long-lasting.

According to PRT, reactance may prompt behavioral resistance wherein individuals attempt to restore and maintain personal autonomy and behavioral freedom by continuing to engage in the restricted behavior, endorsing negative attitudes or opinions about the imposed restrictions, or actively arguing against or seeking out information that refutes the restrictions (for review, see Rosenberg & Siegel, 2018). Persuasive messaging aimed at eliciting behavioral change may also instigate reactance and associated behavioral resistance [13] resulting in a “boomerang effect” wherein message recipients adopt the opposite behavior [14]. In the context of COVID-19, this “boomerang effect” has been reflected in protests against and outright refusal to follow public health guidelines [15, 16]. Burgeoning research suggests that reactance is associated with decreased adherence with social distancing, hand-washing, and mask-wearing [17]. Moreover, reactance to COVID-19 public health measures has been associated with negative attitudes towards masks, increased beliefs about masks being ineffective, and outright refusal to wear a mask [18]. Consequently, utilizing messaging that reduces reactance to public health mandates is a potential mechanism for increasing compliance with these preventive health measures.

Avenues for attenuating reactance: Other-oriented messaging

Reactance is reduced when people are directed to focus on how others, rather than oneself, are affected by restrictions to freedom, known as vicarious reactance [19, 20]. Self-reactance may activate the affective system, leading to impulsive responses, whereas vicarious reactance activates cognitive systems, enabling reflection upon another’s plight with less emotional arousal [21]. Indeed, health messaging that induces empathy or other-oriented thinking is often more persuasive and more successful in eliciting behavioral change as it curbs reactance [14, 22]. Extending this to the current pandemic, public health messaging that focuses attention on how restrictions impact one’s community, rather than oneself, may spur vicarious reactance rather than personal reactance, thereby reducing anger, impulsive actions, and motivation to re-establish freedoms, even among individuals who have been adversely impacted by these restrictions. Consistent with this hypothesis, current public health messaging has adopted a community-focused approach [23], but the extent to which this technique is effective in reducing reactance during the pandemic has yielded conflicting findings.

Specifically, recent research has shown that prosocial messaging with a focus on community, compared to focusing on the self, has had mixed effects on COVID-19 prevention behavior with U.S. samples (for review, see [23]). Some studies have found that messaging that focused on the community increased intentions to wear a mask [24] as well as increased prevention intentions (e.g., handwashing; [25]). However, findings have not been consistent across health behaviors, as message framing that focused on community did not have similar effects for social distancing and understanding COVID-19 through official communication [23]. Similarly, neither Favero and Pedersen [26] nor Ma and Miller [27] found significant differences between self-focused and empathy-inducing or other-focused public health messages on social distancing intentions and other COVID-19 attitudes and beliefs.

One possible explanation for these inconsistent findings is that other factors may influence, or moderate, the impact of public health messaging [27]. That is, contextual factors specific to or resulting from the COVID-19 pandemic and the associated changes to daily life may influence how receptive an individual is towards the framing of pandemic public health messaging. Indeed, Luttrell and Petty [28] found that other-focused public health messages were rated as more persuasive than self-focused messages but only for participants who strongly moralized public health. The present study investigated whether individuals’ experiences of acute, pandemic-related financial stress influenced how the framing of regulations as affecting the community or the self predicted reactance to ongoing and future regulations.

Pandemic-specific financial stress as a moderating influence

COVID-19 not only introduced a national health threat but an economic threat as well [29, 30]. With business closures and workhour reductions, 70% of U.S. families experienced detrimental work impacts by April 2020 [31] leading to greater financial stress [32] and decreased ability to afford basic needs [12]. Thus, public health regulations that limited work activities may have been perceived as personally threatening. Media coverage of hoarding goods and supply chain disruptions may have exacerbated perceptions of threat and need to prioritize self-interests, especially for those experiencing greater financial stress [33]. In addition to the threats to physical health posed by the COVID-19 virus itself, the economic repercussions of the pandemic challenged many people’s financial security.

It is possible that individuals experiencing greater pandemic-specific economic hardship and financial stress may be particularly sensitive and reactive towards messages about pandemic preventative measures, especially those that may infringe upon their livelihood (e.g., non-essential business closure, shelter-in-place). Accordingly, public health regulations that were experienced as both limiting personal agency and jeopardizing financial stability would have been experienced as particularly threatening, and therefore likely to engender greater reactance followed by decreased public health compliance in an effort to reclaim behavioral autonomy. Conversely, public health messaging that orients financially stressed individuals to focus on their community may attenuate reactance as the individual reflects upon pandemic impacts on others in the community.

Critical covariates: Individual differences in pandemic response

Whereas we expected that COVID-related financial stress could moderate the extent to which self- versus community-focused messaging affects reactance, other individual differences could be expected to affect people’s likelihood of reactance to public health measures, regardless of messaging. Three such factors were considered in this study: psychological sense of community (PSOC), political orientation, and gender.

Many people have complied and cooperated with government-mandated or encouraged public health regulations. Those following regulations may feel greater PSOC, or connection to groups impacted by COVID-19 [34, 35]. Greater PSOC is associated with engaging in behaviors that benefit the community while avoiding those that harm [36]. PSOC with others affected by COVID-19 –be they frontline workers, families of infected people, or others–may be expected to curb reactance to loss of freedom through government mandates. Previous research has suggested that cooperation during times of crisis is motivated by shared identity and concern for others [37]. Therefore, PSOC was included as a covariate in this investigation.

In addition to what may facilitate pandemic prevention behaviors, contextual features in the current political climate regarding the emergence, presence, and threats of COVID-19 in the U.S. should be considered. During 2020, in the hyperpolarized political context of the U.S., public health measures were not seen as values-neutral, but as reflective of political orientation [38]. This phenomenon may have been amplified by the ease with which people can select the source of their news, filtering out discrepant messages in favor those confirming one’s position, and thereby adopting or increasing partisan views regarding COVID-19, public health measures and their own health behavior choices [37]. Research on climate change messages suggests that political orientation may be associated with reactance towards public policies based on scientific consensus ([39]; but see [40]). Political orientation may make some individuals less cooperative with public health messaging that contradicts their beliefs about COVID-19, or the actions they see encouraged by the political agents with which they identify. Within the U.S., research on political orientation and pandemic-related attitudes and behaviors suggests Republicans and more politically conservative people perceive less health risk due to COVID-19, whereas Democrats and more politically liberal people are more compliant with public health directives [41, 42]. Consequently, we incorporated political orientation as a second covariate in this investigation.

Finally, recent research into COVID-19 has highlighted gender differences in public health compliance. In general, men tend to engage in fewer preventive behaviors to protect health [25] and more health risk behaviors [43] as compared to women, which may contribute to decreased COVID-19 public health compliance in men. In early June 2020, when mask wearing was recommended by public health officials but prior to widespread mask mandates, women had increased odds (1.5x) of voluntarily wearing masks to a store as compared to men [2]. Similarly, Capraro and Barcelo [24] found that men had less intention to voluntarily wear a face mask and more negative feelings about wearing a mask as compared to women. Recent meta-analytic evidence suggests that rates of mask wearing have been comparable for men and women, but men are more likely to perceive masks as an infringement on their personal freedom [44]; hence, men may have increased reactance toward COVID-19 public health mandates and recommendations. Therefore, we included gender as an additional covariate in our analyses.

Current study

We examined whether self- versus community-focus of restrictive COVID-19 public health messaging affected reactance, and how pandemic-related financial stress was associated with reactance against public health regulations. Additionally, we examined the extent to which reactance towards COVID-19 public health restrictions was associated with social distancing behavior. These effects were explored using an online study between May 7 and June 15, 2020, after widespread U.S. shelter-in-place orders had been established and maintained for several weeks, while controlling for pre-pandemic income and individual differences shown to be associated with pandemic response–specifically, PSOC, political orientation, and gender. We hypothesized that individuals reading self-focused messages about public health restrictions would report greater reactance compared to individuals reading community-focused messages (H1), that pandemic-related financial stress would predict greater reactance (H2), and that the association between financial stress and reactance would be reduced for those in the community-focused condition compared to those in the self-focused condition (H3). Finally, we hypothesized that greater reactance would be associated with less adherence to social distancing guidelines (H4).

Method

Participants

Recruitment was based upon convenience sampling of adults (18 years or older) with access to a computer with internet capabilities. From April 30 to June 15, 2020, advertisements for “an online study to better understand people’s emotional and behavioral responses to the COVID-19 pandemic and the associated public health regulations” were placed on Internet networking sites such as Facebook and disseminated through UC Davis venues. A power analysis to determine an adequate sample size was determined using G*Power for multiple linear regression (R2 increase) testing three predictors in a six predictor model. Parameters were set at an alpha of .05, effect size of .10, and power of .90, which resulted in a minimum of 146 participants to achieve this effect. A total of 386 adults consented to participate (Age; M = 35.74, SD = 14.78, Range: 18–87, 76.04% female, 69.71% White); however, a subset of 301 adults were used in the subsequent analyses (Age; M = 35.88, SD = 14.34, Range: 18–87; see data cleaning for detailed subsetting process). The sample used for subsequent analyses primarily identified as White (n = 218, 72.91%) and female (n = 232, 77.08%) from 17 states and Washington, D.C. (77% California, 4% Colorado, less than 4% for any other state, 4% declined to report). The sample used for subsequent analyses was diverse in their reported income levels, ranging from less than $20,000 to more than $175,000 with each category containing at least 18 participants. However, the largest grouping was between $20,000 - $80,000 (three intervals; n = 125, 41.52%).

Procedure

The UC Davis IRB Administration approved the following project under IRB ID 1587525–1 and determined the project as Exempt 2 status. Consent was obtained from all participants prior to participation. Participants completed the study using Qualtrics, an online data collection tool. Public access to the survey began one week after the first online advertisements, with data collected from May 7, 2020 through June 15, 2020. Prior to completing the survey, participants viewed an online informed consent page. After providing consent, participants completed questionnaires in the following order: demographics (including gender and income), financial stress, adherence to social distancing, political orientation, and psychological sense of community. Participants completed an attention check midway through the survey. Next, Qualtrics randomly assigned participants to read one of two restriction of freedom scenarios, self-focused restriction or community-focused restriction. After reading the scenario, participants answered questions regarding their reactions and responses to what they just read. Participants were thanked for their time and participation at the end of the survey. There was no financial incentive for participation. The survey took participants approximately 15–20 minutes to complete.

Measures

Restriction of freedom scenarios

Participants were randomly assigned to one of two conditions: self-focused restriction or community-focused restriction. Each condition involved reading a prompt describing COVID-19 public health regulations (e.g., social distancing, closure of non-essential businesses) and asking the participant to imagine how these restrictions will continue to impact freedom. In the self-focused condition, participants read a prompt describing how the COVID-19 public health regulations impacted themselves and asking the participants to imagine how these restrictions would continue to impact their own freedom to work, have a social life, and go about daily life (Fig 1A). In the community-focused condition, participants read a prompt describing how the COVID-19 public health regulations impacted others in the surrounding community and asking the participants to imagine how these restrictions would continue to impact community members’ freedom to work, have a social life, and go about daily life (Fig 1B).

Fig 1

Vignettes describing the impact of COVID-19 public health restrictions.

Participants were randomly assigned to be in either the self-focused restriction condition (A) or in the community-focused restriction condition (B).

State reactance

We adapted the Salzburger State Reactance scale [45] to measure participants’ reactance to the restriction of freedom scenarios. The scale consisted of 11 items assessing experience of reactance (e.g., “To what extent do you perceive these actions of the government as a restriction of freedom?”), aggressive behavioral intentions (e.g., “Do you think people should be ignoring, resisting, or protesting these restrictions, limitations, and guidelines?”), and negative attitudes (e.g., “How likely do you think it is that the government will use these actions to take advantage of people?”). Participants answered questions using a 5-point Likert-type scale (1 = Not at all, 5 = Very Much). Mean scores for all items were used to index participants’ state reactance to the restriction of freedom scenarios. Cronbach’s alpha = .90 with bootstrapped 95% CI [.87, .92] based on 1,000 iterations.

Adherence to social distancing

Participants’ adherence to public health guidelines for social distancing was measured with their responses to the item, "Since the pandemic began, I have tried my best to socially distance from others to decrease viral infection in the community". Participants were asked to rate the extent to which they agreed with the statement using a 7-point Likert-type scale (1 = Strongly Disagree, 7 = Strongly Agree). This item was created by the researchers for the purposes of the study.

Financial stress

Participants’ financial stress was measured using 6 items derived from Conger and colleagues’ economic hardship measures [46, 47]. The prompt “Since the pandemic began” was added to items to orient participants to rating financial stress experienced in the wake of the pandemic. The measure included two items assessing inability to make ends meet (e.g., “Since the pandemic began, you have been late paying your bills or have not been able to pay your bills”), two items assessing unmet material needs (e.g., “Since the pandemic began, you have not had enough money to afford the kind of food you needed”), and two items assessing financial strain (e.g., “Your financial situation is worse now than it was before the pandemic began”). Participants were asked to rate the extent to which each statement was true for their circumstances using a 5-point Likert-type scale (1 = Never; not at all true, 5 = Almost every day, definitely true). Mean scores represented participants’ overall financial stress. Cronbach’s alpha = .85 with bootstrapped 95% CI [.79, .88] based on 1,000 iterations.

Gender

Participants’ gender was measured as their response to the following question, “What gender do you identify with?”. Response options included “female”, “male”, “nonbinary or genderqueer”, “transgender”, “other–please specify”, and “decline to answer”. Of the 301 subsetted participants used for analyses, four identified as nonbinary or genderqueer, one identified as other, and one declined to answer. Given the unequal numbers in gender identities, we re-coded responses into a binary variable with 1 indicating female and 0 indicating not female. The one participant who declined to answer was marked as missing data.

Political orientation

Participants’ political orientation was assessed with their responses to the question, “In general, how would you describe your overall political orientation?”. Using a 7-point Likert-type scale, participants were asked to indicate the response that best described them (1 = Very Liberal, 4 = Moderate, 7 = Very Conservative). Higher scores represented more conservative political leaning.

Psychological sense of community

We adapted four items from a measure of psychological sense of community measuring connection and identification [36] to assess participants’ psychological sense of community with those affected by the pandemic (PSOC). Two items measured participants’ connection to those impacted by COVID-19 (e.g., “I feel strong ties to the community affected by COVID-19”) and two additional items measured participants’ personal identification with those impacted by COVID-19 (e.g., “I identify with the community affected by COVID-19”). Participants’ indicated the extent to which they agreed with each statement using a 5-point Likert-type scale (1 = Completely Disagree; 5 = Completely Agree). Mean scores for all four items represented participants’ overall COVID-19 PSOC. Cronbach’s alpha = .93 with bootstrapped 95% CI [.91, .95] based on 1,000 iterations.

Income

Gross household income in 2019 was measured using 9 intervals starting at “Less than $20,000”, “$20,000 to $40,000”, and so on up to “More than $175,000.” Participants were additionally given an option “Don’t know/Decline to answer.” Given the primary focus on financial stress, income was included as another covariate in the tests of hypotheses.

Analytic plan

All analyses were performed in R 1.3.1093. Preliminary analyses were performed with the psych [48] and jmv packages [49]. Path analyses were performed with the lavaan package [50] with simple slopes performed with the semTools package [51].

Data cleaning

The data cleaning procedure included handling missing data as well as removal of participants who did not complete the restriction manipulation, reactance measure, or those who failed the attention check. Of those who consented to participate, 45 discontinued before receiving the restriction manipulation, 7 discontinued without completing the reactance measure, and an additional 5 failed the attention check, which reduced the sample to 329 participants. Finally, 28 participants were removed on the income measure for not knowing or declining to report their income (i.e., selected “don’t know/decline to answer” as their response), which resulted in a final data set for analyses of 301 adults. On the variables of interest, Little’s MCAR test was conducted, χ2(5) = 11.06, p = .05, and confirmed the data was missing completely at random. Of note, the only variable with missing data was financial stress (percentage missing 1.96%). Thus, given the small amount of missing data, full information maximum likelihood (FIML) was used during path analysis to account for missingness.

Assumptions

The analysis plan included using path analysis with a binary-continuous interaction term between financial stress and restriction condition. The multivariate normality assumption for path analysis regression was not met as demonstrated by the significant Mardia’s skewness (p < .001) and kurtosis (p < .001; [52]). Given the violation of multivariate normality, we used Maximum Likelihood with robust corrections to fit the model and obtain parameter estimates with robust standard errors [53, 54].

Open practices statement

The current analysis was not formally preregistered. The data and full reproducible code can be accessed via https://github.com/meknapp18/ProReact.

Results

Descriptive statistics

Table 1 contains zero-order correlations by restriction condition for all key study variables, as well as means, standard deviations, skew, and kurtosis. For the entire sample, state reactance was positively correlated with financial stress (r = .19, p = .001) and political orientation (r = .36, p = .034) and negatively correlated with PSOC (r = -.28, p < .001), income (r = -.12, p = .04), and adherence to social distancing (r = -0.42, p < .001). State reactance was not significantly correlated with gender (r = .02, p = .71). For the self-focused condition (n = 148), state reactance was positively correlated with financial stress and political orientation, was negatively correlated with PSOC and with adherence to social distancing, and was not associated with income or with gender. For the community-focused condition (n = 153), state reactance was positively correlated with political orientation, negatively correlated with PSOC and with adherence to social distancing, and not associated with either financial stress, income, or gender.

Table 1

Descriptive statistics and zero-order correlations for reactance and predictors.

Variables	1	2	3	4	5	6	7
1. Reactance	—	−.36***	.30***	.03	.37***	−.29***	−.13
2. Adherence to Social Distancing	−.50***	—	−.21**	−.01	.23**	.26**	.18*
3. Financial Stress	.08	−.22**	—	.02	.02	.01	−.43***
4. Gender	−.07	.18*	.04	—	−.01	−.12	.08
5. Political Orientation	.34***	−.31***	.07	−.01	—	−.09	.15t
6. PSOC	−.29***	.11	−.10	−.01	−.01	—	−.01
7. Income	−.11	.13	−.23***	.07	.04	.16*	—
M	2.32	6.30	1.48	0.80	2.78	3.26	4.91
SD	0.80	1.07	0.72	0.40	1.51	1.01	2.60
Skew	1.15	−2.28	2.17	−1.48	0.83	−0.53	0.23
Kurtosis	1.32	6.40	4.65	0.20	0.14	−0.17	−1.16

N = 301. Zero-order correlations for the self-focused condition (n = 148) are reported above the diagonal. Zero-order correlations for the community-focused condition (n = 153) are reported below the diagonal. PSOC = Psychological sense of community. Gender coded 1 = Female, 0 = Not Female.

p < .10,

* p < .05,

** p < .01,

*** p < .001.

As a check on the randomization process, we performed planned comparisons with between-subjects’ t-tests to probe for differences between restriction conditions on financial stress, PSOC, political orientation, income, gender, and adherence to social distancing. Those in the community-focused condition were more politically conservative (M = 2.99, SD = 1.46) compared to those in the self-focused condition, M = 2.57, SD = 1.53, t(299) = -2.44, p = .015, 95% CI: [-0.76, -0.08]. There were no other significant differences between restriction conditions on predictors and covariates (p’s > .05).

Path analysis

A path model included direct pathways from all exogenous study variables (e.g., financial stress, dummy-coded restriction condition, etc.) and the interaction of financial stress and restriction condition to state reactance. In addition, based on the correlation matrix, covariance between restriction condition and political orientation, financial stress and income, and the interaction term and its components were specified. Both standardized (Table 2) and unstandardized path coefficients (Fig 2) were estimated, resulting in model fit indices of: χ2(24) = 44.03, p = .01, robust CFI = .97, robust TLI = .95, robust RMSEA = .05 (90% CI: [.03, .08]), and robust SRMR = .06. The likelihood-ratio test was significant, suggesting that our model-implied covariance matrix was not equal to that of the population; that is, our model was not “true” or exact to the population [55]. However, path analyses with large sample sizes provide the smallest confidence intervals around parameter estimates, resulting in the likelihood-ratio test becoming overpowered and more likely to reject models due to small differences in residuals when the test model is well-specified [56].

Table 2

Maximum-likelihood robust models of financial stress, restriction condition, and their interaction predicting reactance to COVID-19 public health restrictions.

Effect	β (SE)	z-value
Predictors
Reactance → Adherence to Social Distancing	−.42 (.11)	−5.22***
Financial Stress → Reactance	.26 (.10)	2.88**
Restriction Condition → Reactance	.22 (.20)	1.72t
Financial Stress*Restriction Condition → Reactance	−.32 (.13)	−2.19*
Covariates
PSOC →Reactance	−.27 (.05)	−4.35***
Political Orientation →Reactance	.35 (.04)	5.24***
Income →Reactance	−.07 (.02)	−1.29
Gender →Reactance	−.04 (.10)	−0.67
Covariances
Restriction Condition ↔ Political Orientation	.02 (.02)	0.85
Restriction Condition ↔ Interaction	.83 (.01)	35.39***
Financial Stress ↔ Income	−.23 (.08)	−5.37***
Financial Stress ↔ Interaction	.36 (.05)	4.82***

N = 301. PSOC = Psychological sense of community.

< .10,

* p < .05,

** p < .01,

*** p < .001.

Fig 2

Path model predicting adherence to social distancing guidelines and reactance to COVID-19 public health restrictions from financial stress, restriction condition (self- or community-focused), and the interaction between financial stress and restriction condition while controlling for gender, psychological sense of community, political orientation, and income.

Values are unstandardized regression estimates with standard errors in parentheses. * p < .05, ** p < .01, *** p < .001.

As a robustness check, we performed the RMSEA test of close fit (H: model has good fit) and the test of not-close fit (H: model has poor fit) as outlined by MacCallum and colleagues [57]. For the RMSEA fit index, the asymptotic sampling distribution is known [58, 59], which affords the ability to compute confidence intervals and to conduct null hypothesis significance testing (NHST) about the RMSEA population value. For both the test of close fit and not-close fit, we computed the noncentrality parameter that corresponds to a specified RMSEA cutoff value followed by a one-tailed NHST using the non-central chi-square distribution. For the test of close fit, we specified an RMSEA cutoff value of .05 with NHST results suggesting close model fit as we failed to reject the null hypothesis, χ2(24) = 61.38, p > .05. For the test of not-close fit, we specified an RMSEA cutoff value of .11 with NHST results suggesting that the model does not have poor fit as we reject the null hypothesis, χ2(24) = 80.08, p < .05. Consequently, these results combined with the other comparative and absolute fit indices indicate that our model has good fit despite the significant likelihood-ratio test. Although restriction condition did not significantly predict reactance, b = .35, p = .09, 95% CI [-0.05, 0.75], there was a significant main effect of financial stress (b = .30, p = .004, 95% CI [0.09, 0.50]), and the interaction of financial stress with restriction condition was significant (b = -.29, p =. 03, 95% CI [-0.55, -0.03]). For the two-way interaction, simple slopes analysis revealed that financial stress was positively associated with state reactance for those in the self-focused condition, b = 0.29, SE = .10, p = .004. However, for those in the community-focused condition, there was not a significant association between financial stress and reactance, b = 0.01, SE = .09, p = .96. To contextualize the interaction, we performed a second simple slopes analysis with financial stress moderating restriction condition, allowing us to probe for differences between restriction condition at the mean level of financial stress as well as one standard deviation above and below the mean. Among individuals experiencing high levels of financial stress (+1 SD = 2.21), being in the community-focused condition predicted less reactance than being in the self-focused condition (b = -0.29, SE = .13, p = .03), but reactance did not differ significantly between conditions at mean (M = 1.49, b = -0.08, SE = .08 p = .30) or low levels of financial stress (-1 SD = .08, b = 0.12, SE = .12 p = .27; Fig 3).

Fig 3

Simple slopes analyses for the interaction between restriction condition (self- and community-focused) and financial stress predicting state reactance to COVID-19 public health restrictions, controlling for gender, psychological sense of community, political orientation, and income.

Financial stress was positively associated with state reactance for those in the self-focused condition but was nonsignificant in the community-focused condition.

For covariates, PSOC significantly predicted decreased reactance (b = -.21, p < .001 95% C1 [-0.30, -0.11]) and political orientation significantly predicted increased reactance (b = .19, p < .001, 95% CI [0.12, 0.26]). Recall that a higher score for political orientation represents a more conservative political orientation, suggesting that more conservative individuals had greater reactance. Income (b = -0.02, p = .20, 95% C1 [-0.06, 0.01]) and gender (b = -0.07, p = .50, 95% CI [-0.13, 0.26]) did not predict reactance.

Finally, reactance was significantly associated with adherence to social distancing guidelines, b = -0.57, p < .001, 95% CI [-0.78, -0.35]. That is, those reporting greater reactance also reported lower levels of social distancing behavior. Overall, the variance explained by the model for state reactance was R2 = .25 and for adherence to social distancing was R2 = .18 with an adjusted GFI = .99 for the entire model. This suggests that 25% of the variance in state reactance, reflecting a large effect size, 18% of the variance in adherence to social distancing, reflecting a medium-to-large effect size, and that 99% of the variance in the sample covariance matrix were accounted by the model.

Exploratory post hoc analysis

Given that, despite random assignment to condition, more politically conservative individuals were over-represented in the community-focused condition, we examined a second, exploratory path model that included a restriction condition x political orientation interaction effect. The exploratory path model had poor fit, χ2(30) = 274.31, p < .001, robust CFI = .74, robust TLI = .61, robust RMSEA = .19 (90% CI: [.17, .21), and robust SRMR = .13. Moreover, the restriction condition x political orientation interaction effect was not significant, b = -0.01, p = .85, 95% CI [-0.15, 0.13], and including it in the model did not change the nonsignificant main effect for restriction condition, b = .39, p = .09, 95% CI [-0.05, 0.83] nor the significant restriction condition x financial stress interaction effect, b = -0.29, p = .03, 95% CI [-0.56, -0.03].

Discussion

Soon after governments in the U.S. began issuing public health guidelines and regulations to stem the transmission of COVID-19 in the early spring of 2020, people began complaining, defying, and protesting against these measures [16]. The current study demonstrates that adults who experienced more pandemic-related financial stress were more likely to react against restrictive public health measures. In accordance with the vicarious reactance tenet of psychological reactance theory (PRT), however, public health messaging that drew focus to the effects of such measures on the community, rather than the self, attenuated financially-stressed adults’ reactance. These effects were independent of adults’ actual income and gender, and over-and-above the tendencies for more conservative individuals to report greater reactance against regulations, and those who felt more connected with COVID-affected groups (PSOC) to report less reactance. Notably, greater reactance towards COVID-19 public health messaging was associated with lower adherence with social distancing guidelines, a critical preventive health behavior in the pandemic.

These findings have implications both for understanding why some individuals have actively resisted cooperating with public health regulations and for developing more effective messaging to solicit their greater cooperation in future. Moreover, the rapid spread and climbing infection rate from the SARS-CoV-2 B.1.617.2 variant (i.e., the delta variant) has instigated worldwide discussions of implementing COVID-19 vaccine mandates [60] and re-instating mask mandates even for those who are vaccinated [61]. Emerging research suggests that vaccine mandates for adults will likely prompt anger, heightened reactance, and defiance and backlash towards such mandates [62, 63]. Our findings build upon current COVID-19 health messaging research by underscoring the need to consider the contextual factors specific to COVID-19 in combination with message framing to better tailor public health messaging, decrease reactance towards mandates, and elicit greater compliance with public health guidelines.

The lack of a main effect of restriction condition suggests that, contrary to previous PRT research [20] and our hypotheses, the intensity of personal and vicarious reactance did not differ significantly. However, our findings are consistent with recent research that has also failed to find differences between self- and other-focused messages in reactance [27] or in COVID-19 attitudes and beliefs [26]. It is possible that the COVID-19 pandemic and associated public health measures had been so pervasive and such drastic deviations from everyday life in the U.S. that the framing of messaging, in and of itself, did not influence people’s negative responses to continued freedom loss. Some previous investigations of self-focused and vicarious reactance also failed to find differences between restriction conditions on state reactance [27, 45], or differences that were significant only in the context of moderator variables [21]. Such findings suggest that differences between self-focused and vicarious reactance could depend upon other factors, such as pandemic-related financial stress as observed in the present study. Additionally, these findings in addition to others, highlight the complexity of responses to the pandemic, such that messages which imply that people have agency in spreading the virus, rather than lower control, may lead people to having less reactive emotions to restrictions [27]. Overall, message framing and the intricacies of the source, target, among other factors are worthy of consideration when attempting to understand responses to public health messaging.

The vignettes asked participants to reflect upon how restrictions would continue to impact the self or community in the future, which may have prompted anxiety or anger about limitations and hardships continuing indefinitely for people who, given other studies, likely were already feeling stressed [64]. Clearly, imagining continued employment disruptions elicited considerable negativity and defiance in those currently feeling financial anxiety about making ends meet, yet their reactance was diminished when their attention was focused on how continued public health measures would affect the community. The vicarious reactance condition should have led participants to adopt a more cognitive, reflective approach to considering the situation, rather than an emotional, impulsive stance [5, 21]. Thus, reminders that “we are all in this together” may lessen reactance in financially stressed individuals as they contemplate how others may be equally or more impacted by public health measures, and conversely, how their own situation may not be as uniquely difficult as imagined. Looking toward the future and the possibilities for continuing public health measures in response to COVID-19 variant strains, or other yet-to-be identified pandemics, we propose that community-focused messaging regarding individual and societal behavioral changes may help to curb reactance in those who feel personally adversely affected, which may increase cooperation and the success of public health efforts.

Notably, PSOC and political orientation both significantly predicted reactance, elucidating the importance of individual differences in predicting responses to pandemic conditions. Lower reactance in those with greater PSOC with affected groups reemphasizes the importance of community-focused messaging and community-building as an effective public health strategy [23, 24, 65]. Greater reactance in more politically conservative adults mirrors the increasingly polarized U.S. political landscape [66] which has influenced how people perceive and comply with public health regulations and messages. People are more receptive towards messages that align with their attitudes [67, 68], and more reactant towards messages that challenge their identity or beliefs [69, 70]. Further research is warranted to better understand the role of political orientation in public health messaging and compliance with regulations, but emphasizing the benefits of public health measures for the communities and priorities with which more conservative adults identify may be one way to enlist their cooperation.

Strengths and limitations

Our study had many strengths. We measured peoples’ reactance soon after widespread implementation of COVID-19 emergency orders, revealing how the public acutely responded to an unprecedented global health crisis; we examined responses in a large, geographically diverse sample; and the randomized manipulation of restriction scenarios provided evidence that focusing on one’s community produces lower reactance in financially-stressed adults. Our findings expand upon the growing literature exploring the impact of prosocial, other-oriented appeals for public health compliance [23, 37], providing nuance in demonstrating reactance and pandemic-specific stressors as key processes underlying associations between message framing and health behavior. With increasing uncertainty around emerging variants, vaccine booster shots, and a possible return to mask mandates, our findings provide critical insight into the direct and interactive influences of message framing and pandemic-specific stressors in dampening the “boomerang effect” and promoting health behavior compliance as the pandemic continues.

Of note, our study measured individual’s reported reactance to COVID-19 public health restriction scenarios and the extent to which they tried to comply with social distancing; we did not ask participants to report their actual compliance with COVID-19 public health regulations or their perceived risk of COVID-19. Additionally, we focused on social distancing behavior exclusively as mask wearing had not been mandated and was actively cautioned against at the time of study design [71, 72]. Thus, we have a limited understanding of how financial stress, message framing, and reactance uniquely and jointly predict other types of pandemic public health behaviors, such as wearing a mask or receiving the vaccine.

Additionally, we measured a limited number of individual difference variables. Emerging research suggests that personality traits are differentially associated with pandemic response and COVID-19 health behaviors [73]. Following a socioeconomic determinants of health perspective [74], an individual’s occupation (e.g., frontline worker, essential worker) may also influence the individual’s attitudes and beliefs towards COVID-19, the associated public health mandates, and compliance with preventative health behavior. We did not collect data on personality and occupational status, but future research should explore how these additional individual difference factors relate to public health message framing, reactance, and COVID-19 public health compliance.

Despite random assignment to restriction condition, the disproportionate representation of more politically conservative participants within the community-focused group may have elevated reactance within this condition and reduced the magnitude of the predicted difference between restriction conditions; future research should explore this possibility further. Additionally, there was a disproportionate distribution of gender in our study. The predominance of females in our sample limits the generalizability of our findings to other genders. The majority of research on gender effects in response to COVID-19 has found greater outrage, reactance, and public health noncompliance in males [2, 24, 44], but this was not evident in our study. It behooves future research to further explore how public health message framing influences reactance in individuals who do not identify as female.

Finally, the 2020 U.S. federal election campaign and the nationwide community actions sparked by the murder of George Floyd on May 25, 2020, overlapped with our data collection; either may have affected how participants engaged with questions pertaining to government mandates. Additionally, our sample was predominantly White. Given that COVID-19 has disproportionately impacted Black and Latinx communities [75, 76], concerted efforts are needed to better understand how to best support these communities, including whether different approaches to public health messaging may be needed to more effectively reduce their risks.

Conclusion

Community-focused messaging reduced reactance to public health regulations restricting activities in adults already feeling financially stressed due to efforts to curb transmission of the COVID-19 pandemic. This suggests that careful tailoring of public health messages may be effective for increasing compliance and cooperation, or at least limiting the harm of pushing against regulations. Ultimately, the success of the nation in appealing to citizens regarding preventative health behavior and restrictions is partly due to the way regulations are tailored. The economic, social, and personal effects of the pandemic have been tremendous, and at times overwhelming, for many people, yet shifting the frame of their focus towards the community may limit some of the emotional impact of ongoing turmoil due to the pandemic.

21 Jul 2021

PONE-D-21-19560

We’re all in this together: Focus on community attenuates effects of pandemic-related financial hardship on reactance to COVID-19 public health regulations

PLOS ONE

Dear Dr. Knapp,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please find below the reviewer's comments, as well as those of mine.

Please submit your revised manuscript by Sep 04 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Valerio Capraro

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please ensure that you include a title page within your main document. We do appreciate that you have a title page document uploaded as a separate file, however, as per our author guidelines (http://journals.plos.org/plosone/s/submission-guidelines#loc-title-page) we do require this to be part of the manuscript file itself and not uploaded separately. Could you therefore please include the title page into the beginning of your manuscript file itself, listing all authors and affiliations.

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Additional Editor Comments (if provided):

I have now collected one review from one expert in the field. I was unable to find a second reviewer, but I am myself familiar with the topic of this manuscript, so I feel confident in making a decision with only one review. As you will see, the review is positive, but suggests a major revision. I tend to agree with the reviewer, therefore I would like to invite you to revise your manuscript for Plos One. On top of the reviewer's comments, I would like to add two more comments, mainly related to a potential effect of gender, and to the literature review. Indeed, I have noticed that in your sample there are many more females than males. Since there are gender differences in pandemic response (Capraro & Barcelo, 2020; Haischer et al. 2020), it is important to control for gender (it seems to me that you did not do so). Regarding the literature review, I was surprised to see that you ignored the previous papers on the effect of prosocial, community-oriented, messages on pandemic response, which seem to be directly related to your work (Capraro & Barcelo, 2020; Jordan et al., 2020). More generally, there has been a number of papers looking at the effect of prosocial messages on pandemic response (see Capraro et al. 2021 for a review). Another work that you might find useful is the perspective article on what social and behavioral science can do to support pandemic response published by Van Bavel et al. in Nature Human Behaviour.

I am looking forward for the revision.

Capraro, V., & Barcelo, H. (2020). The effect of messaging and gender on intentions to wear a face covering to slow down COVID-19 transmission. Journal of Behavioral Economics for Policy, 4, Special Issue 2, 45-55.

Capraro, V., Boggio, P.S., Böhm, R., Perc, M., & Sjåstad, H. (forthcoming) Cooperation and acting for the greater good during the COVID-19 pandemic. In M. K. Miller (Ed.) The

Social Science of the COVID-19 Pandemic: A Call to Action for Researchers. Oxford: Oxford University Press. Available at: https://doi.org/10.31234/osf.io/65xmg

Haischer, M. H., Beilfuss, R., Hart, M. R., Opielinski, L., Wrucke, D., Zirgaitis, G., ... & Hunter, S. K. (2020). Who is wearing a mask? Gender-, age-, and location-related differences during the COVID-19 pandemic. Plos one, 15(10), e0240785.

Jordan, J., Yoeli, E., & Rand, D. (2020). Don’t get it or don’t spread it? Comparing self-interested versus prosocially framed COVID-19 prevention messaging. https://psyarxiv.com/yuq7x

Van Bavel, J. J., et al. (2020). Using social and behavioural science to support COVID-19 pandemic response. Nature Human Behaviour, 4, 460-471.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you very much for the opportunity to revise this manuscript. I believe the manuscript has the potential to add to existing research on compliance with COVID-19 Public Health Regulations. Below, I state my comments that I hope you would find it useful.

1- I appreciate the authors’ introduction in terms of the contribution of the psychological reactance theory to this literature. However, all of a sudden, the manuscript started to present thoughts about republicans, and democrats in page 5. There is a lack of flow in the text in this part. The political orientation in fact seems very disconnected from the basic premises of the work.

2- The way I read the empirical arguments is that Finanical stress results in reactance towards COVID-19 public health restrictions, which in turn affect social distancing behavior. Individual differences in PSOC, and political orientation moderate these associations. Yet, the hypotheses do not explain at which step do the moderators take place. Neither does the model in Figure 2.

3- The replicability link does not work, showing that the depository is empty.

4- Further descriptives are needed, such as the distribution of the main variables.

5- I wonder what is the reason for using path analysis for the direct effects? A general OLS for the direct effects and give a better glimpse about the effect sizes.

6- I wonder if the authors could address causality issue in a way or another, specifically, could it be that adherence to social distancing are the one that resulted in financial stress, specially that social distancing rules are the one that could be considered an exogenous shock in this case.

7- Further controls are needed, such as personality traits (i.e. research has shown that stable personality traits such as the big five, and optimism as well as occupations might play a role in compliance with COVD-19 health regulations.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

30 Aug 2021

We thank the anonymous Editor and Reviewer for the thorough and helpful reviews they provided to the original manuscript. We have endeavored to respond to each of the comments raised in the revised manuscript and in this response letter. The specific ways in which we have responded to each of the comments is detailed below with our responses provided in italicized font.

Additional Editor Comments (if provided):

I have now collected one review from one expert in the field. I was unable to find a second reviewer, but I am myself familiar with the topic of this manuscript, so I feel confident in making a decision with only one review. As you will see, the review is positive, but suggests a major revision. I tend to agree with the reviewer, therefore I would like to invite you to revise your manuscript for Plos One. On top of the reviewer's comments, I would like to add two more comments, mainly related to a potential effect of gender, and to the literature review.

Indeed, I have noticed that in your sample there are many more females than males. Since there are gender differences in pandemic response (Capraro & Barcelo, 2020; Haischer et al. 2020), it is important to control for gender (it seems to me that you did not do so).

We originally excluded gender as a covariate as preliminary analyses indicated that it was not significantly correlated with any of the key study variables nor were there any mean-level differences between gender for any of the key study variables. However, as you have noted, the COVID-19 literature has found notable gender differences in pandemic response. In the introduction, we have included a paragraph describing gender differences in pandemic response and how it may relate to reactance (p. 5-6). Additionally, we have re-run our path analysis to include gender as a covariate and have updated the Results and associated Tables and Figures to reflect the new model. Additionally, we are heartened that key findings were maintained in the new model controlling for gender.

Regarding the literature review, I was surprised to see that you ignored the previous papers on the effect of prosocial, community-oriented, messages on pandemic response, which seem to be directly related to your work (Capraro & Barcelo, 2020; Jordan et al., 2020). More generally, there has been a number of papers looking at the effect of prosocial messages on pandemic response (see Capraro et al. 2021 for a review). Another work that you might find useful is the perspective article on what social and behavioral science can do to support pandemic response published by Van Bavel et al. in Nature Human Behaviour.

I am looking forward for the revision.

Capraro, V., & Barcelo, H. (2020). The effect of messaging and gender on intentions to wear a face covering to slow down COVID-19 transmission. Journal of Behavioral Economics for Policy, 4, Special Issue 2, 45-55.

Capraro, V., Boggio, P.S., Böhm, R., Perc, M., & Sjåstad, H. (forthcoming) Cooperation and acting for the greater good during the COVID-19 pandemic. In M. K. Miller (Ed.) The Social Science of the COVID-19 Pandemic: A Call to Action for Researchers. Oxford: Oxford University Press. https://doi.org/10.31234/osf.io/65xmg

Haischer, M. H., Beilfuss, R., Hart, M. R., Opielinski, L., Wrucke, D., Zirgaitis, G., ... & Hunter, S. K. (2020). Who is wearing a mask? Gender-, age-, and location-related differences during the COVID-19 pandemic. Plos one, 15(10), e0240785.

Jordan, J., Yoeli, E., & Rand, D. (2020). Don’t get it or don’t spread it? Comparing self-interested versus prosocially framed COVID-19 prevention messaging. https://psyarxiv.com/yuq7x

Van Bavel, J. J., et al. (2020). Using social and behavioural science to support COVID-19 pandemic response. Nature Human Behaviour, 4, 460-471.

We appreciate the Editor’s recommended literature and have incorporated the suggested papers into our introduction. Structurally, we have added the following subheadings to the introduction: Reactance in the Pandemic Context (p. 2-4), Avenues for Attenuating Reactance: Other-Oriented Messaging (p. 4-5), Pandemic-Specific Financial Stress as a Moderating Influence (p. 5-6), and Critical Covariates: Individual Differences in Pandemic Response (p. 6-8). We have incorporated your suggested literature into the Avenues for Attenuating Reactance: Other-Oriented Messaging subsection (p. 4-5) where we have expanded upon public health message framing and differential impacts on COVID-19 preventative behavior and public health compliance. Additionally, we have expanded our Discussion to include how our findings relate back to the growing literature regarding prosocial, other-oriented messages in the context of COVID-19 (p. 21-22, 24-25).

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you very much for the opportunity to revise this manuscript. I believe the manuscript has the potential to add to existing research on compliance with COVID-19 Public Health Regulations. Below, I state my comments that I hope you would find it useful.

1- I appreciate the authors’ introduction in terms of the contribution of the psychological reactance theory to this literature. However, all of a sudden, the manuscript started to present thoughts about republicans, and democrats in page 5. There is a lack of flow in the text in this part. The political orientation in fact seems very disconnected from the basic premises of the work.

We appreciate the Reviewer’s comments regarding manuscript flow and cohesion. We have restructured the introduction and have added the following subheadings to the introduction: Reactance in the Pandemic Context (p. 2-4), Avenues for Attenuating Reactance: Other-Oriented Messaging (p. 4-5), Pandemic-Specific Financial Stress as a Moderating Influence (p. 5-6), and Critical Covariates: Individual Differences in Pandemic Response (p. 6-8). We have expanded upon political orientation as a salient individual difference factor in the context of COVID-19, providing further explanation as to how political orientation may relate to reactance and therefore should be controlled for in subsequent analyses (p. 7-8).

2- The way I read the empirical arguments is that Finanical stress results in reactance towards COVID-19 public health restrictions, which in turn affect social distancing behavior. Individual differences in PSOC, and political orientation moderate these associations. Yet, the hypotheses do not explain at which step do the moderators take place. Neither does the model in Figure 2.

We appreciate the Reviewer’s statement of their interpretation of the empirical arguments; however, it was not our intent to suggest that PSOC and political orientation moderate these associations. Rather, our argument is that PSOC and political orientation may be associated with how people respond to COVID-19 public health messaging; thus, they are necessary covariates for analyses aimed at examining the direct and interactive effects of restriction condition and pandemic-specific financial stress. We have re-structured and expanded the introduction to better articulate our empirical arguments and hypotheses. Additionally, we have re-ordered the Measures subsection to present focal measures related to hypotheses first, followed by covariates (p. 11-14). Hopefully, our line of reasoning is conveyed more clearly now.

Additionally, to ensure that we were not erring in treating these measures as covariates rather than moderators, we tested alternative models that included moderating effects of political orientation and PSOC on restriction condition and reactance. Neither interaction term was significant: political orientation x restriction condition (b = -.02, p = .73); PSOC x restriction condition (b = .04, p = .67).

3- The replicability link does not work, showing that the depository is empty.

We apologize for the oversight. It was our understanding that the replicability link could be provided upon acceptance of the manuscript; therefore, it was not provided. We have corrected our error and the replicability link should now work.

4- Further descriptives are needed, such as the distribution of the main variables.

We have added information on skew and kurtosis to Table 1 (p. 12-13) for all key study variables to provide further information about the distribution of the main variables. If other descriptive statistics are needed then we are happy to provide them.

5- I wonder what is the reason for using path analysis for the direct effects? A general OLS for the direct effects and give a better glimpse about the effect sizes.

Although OLS regression may be considered a more straightforward or accessible approach to addressing the direct effects, we opted to use path models in a structural equation modelling framework as this provides numerous quantitative advantages over OLS. To be succinct, SEM affords more model specification flexibility, generates fit indices, can correct for multivariate nonnormality in comparison to OLS, and better handles missing data (Kline, 2011).

To expand on these points, SEM is more flexible and amenable to model specification of hypothesized or known associations between variables whereas OLS does not allow this type of model specification (Kline, 2011). For example, path analysis permits us to model specific covariances between predictors and covariates whereas multiple regression assumes that all covariates and predictors meaningfully covary. SEM also allows us to model predictive paths between multiple endogenous variables in one model, affording a more parsimonious analysis in comparison to multiple OLS models. SEM allows us to impose a specific covariance structure that reflects theory or hypothesized associations between variables, allowing us to then test “goodness-of-fit” of this theoretically-based covariance structure (Guttman, 1971; Hayduk et al., 2007). Additionally, SEM provides us with multiple model fit indices to assess how well our proposed model reflects the covariance in the data and the extent to which our hypothesized associations are plausible in the population whereas OLS only provides effect size of variance explained by the model, but not fit (Kline, 2011; MacCallum & Austin, 2000). OLS assumes multivariate normality, an assumption which our data has violated. Unfortunately, OLS does not afford robust corrections to violations of multivariate normality whereas SEM can correct for this using Maximum Likelihood with Robust Corrections estimation (Mardia, 1971; Satorra & Bentler, 1994). Finally, SEM allowed us to use sophisticated techniques like Full Information Maximum Likelihood for estimating missing data (Kline, 2011).

Yet, as a robustness check, we have examined two hierarchical multiple regression models to test our hypotheses predicting reactance and one linear multiple regression model to test our hypothesis for reactance predicting adherence to social distancing behavior. Please note that multivariate normality was violated in these analyses. The results of the two hierarchical models indicate that the model with the interaction term explains significantly more variance than the model without the interaction term. Moreover, all of our findings from the SEM are maintained in the model with the interaction term. In the final model predicting adherence to social distancing guidelines, we found continued support for our SEM findings with greater reactance significantly predicting decreased adherence to social distancing. The regression models did not identify new associations that were not evident in SEM. We are heartened by the robustness of our findings across analytic approaches, and we continue to utilize and report the SEM analyses because of their considerable quantitative advantages.

We understand the Reviewer’s concern regarding OLS and effect size. We have reported both R2 and GFI but have also added the following verbiage to increase clarity:

“This suggests that 25% of the variance in state reactance, reflecting a large effect size, 18% of the variance in adherence to social distancing, reflecting a medium-to-large effect size, and that 99% of the variance in the sample covariance matrix were accounted by the model.” (p. 20)

6- I wonder if the authors could address causality issue in a way or another, specifically, could it be that adherence to social distancing are the one that resulted in financial stress, specially that social distancing rules are the one that could be considered an exogenous shock in this case.

We understand the Reviewer’s concern regarding alternative models and causality and appreciate the Reviewer’s suggestion of examining the effects of social distancing as an exogeneous shock. Of course, the experimental component of this study was the random assignment of participants to the self-focused versus community-focused restriction condition, and causal inferences should only be drawn about the borderline-significant main effect of restriction condition. Individual variation in financial stress may be seen as a result of the “natural experiment” of COVID-19, but lacking random assignment or pre-COVID measures, caution must be exercised regarding the inference of causal or directional relations between financial stress and reactance against (or adherence with) public health restrictions.

That said, we tested the model that the Reviewer suggested, removing the path in which reactance predicts adherence to social distancing and adding a path in which adherence to social distancing predicts financial stress. The alternative model has good to adequate fit, χ^2 = 57.49, p < .001, robust CFI = .94, robust TLI = .90, robust RMSEA = .08 (90% CI: [.06, .11]), robust SRMR = .07. However, these fit indices are inferior to those in our theorized model. Additionally, the alternative model has a higher AIC and BIC value (AIC = 6537.34, sample-size adjusted BIC = 6555.02) in comparison to our theorized model (AIC = 6515.71, sample-size adjusted BIC = 6531.78). Finally, we calculated an AIC evidence ratio (Burnham et al., 2011) of 49786.54, suggesting that the theorized model is approximately 50,000 times stronger than the alternative model. Of note, adherence to social distancing significantly predicted decreased financial stress (b = -0.08, p = .02). From a theoretical perspective, we would expect this path coefficient to be positive wherein social distancing is contributing to increased financial stress, perhaps by impacting ability to work; however, the quantitative evidence does not suggest this. We are encouraged by these results and will continue to report and interpret the findings from our theorized model. We have not reported the alternate model in the revised manuscript; however, if the Reviewer and Editor think that it is important information to include, we would be happy to include the alternate model in the main text or an online supplement.

We would also like to emphasize that our measure of financial stress was designed to capture a wide-range of economic impacts from the pandemic, some of which would not be theoretically linked to social distancing. We have added more example items from the measure to increase clarity for the reader (p. 12).

7- Further controls are needed, such as personality traits (i.e. research has shown that stable personality traits such as the big five, and optimism as well as occupations might play a role in compliance with COVD-19 health regulations.

We appreciate the Reviewer’s suggestion of controlling for personality traits and work occupation; unfortunately, these variables were not measured in our study and cannot be included as covariates. However, we have included gender as an additional covariate with no impact to our findings. In the Discussion, we have also recommended examination of personality traits and occupational status as avenues for future research.

“Additionally, we measured a limited number of individual difference variables. Emerging research suggests that personality traits are differentially associated with pandemic response and COVID-19 health behaviors (Corr, 2021). Following a socioeconomic determinants of health perspective (Khalatbari-Soltani et al., 2020), an individual’s occupation (e.g., frontline worker, essential worker) may also influence the individual’s attitudes and beliefs towards COVID-19, the associated public health mandates, and compliance with preventative health behavior. We did not collect data on personality and occupational status, but future research should explore how these additional individual difference factors relate to public health message framing, reactance, and COVID-19 public health compliance.” (p. 24-25)

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

12 Oct 2021

PONE-D-21-19560R1We’re all in this together: Focus on community attenuates effects of pandemic-related financial hardship on reactance to COVID-19 public health regulationsPLOS ONE

Dear Dr. Knapp,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. I was able to secure one review from an expert in the field. The reviewer has provided excellent feedback and has highlighted some remaining concerns. Please be sure to address all of these remaining issues in your revision.Of particular importance, the reviewer noted that the replicability link still does not work.

Please submit your revised manuscript by Nov 26 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Neha John-Henderson

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors sufficiently responded to the comments from the Editor and Reviewer. I appreciate the amount of detail and effort that went into each of the responses. In particular, I appreciate the inclusion of gender in the analyses. I also appreciate the inclusion of subheadings in the introduction, as this makes for an easier read.

MAJOR NOTES

-Following up on the original comment from Reviewer 1, the replicability link still does not work, showing that the depository is empty.

-Where did the item for “Adherence to Social Distancing” come from? Is this pulled from another questionnaire, or was it developed by the authors themselves for the purpose of this study? Please state either way.

-Was the wording of the 6 items from Conger and colleagues’ economic hardship measures adapted to include “since the pandemic began”? If so, please state that you adapted/changed this measure to meet the needs of the current study.

-Under “Psychological Sense of Community”, you provide an example item for the two items measuring personal identification, but you do not provide an example item for the two items measuring connection. While not a huge deal, I recommend also including an example item of connection for your readers.

-You don’t mention income as a covariate in your introduction, but you mention it in your method section, and even use it as an exclusion criterion if not answered. Perhaps briefly mention it in your “Current Study” section of the introduction.

-Table 1 is very helpful, and I appreciate its inclusion in the manuscript. However, it does not follow APA formatting guidelines. Additionally, many of the values do not line up, such that some are flush left, and some are centered, making it harder to read. Please reformat this table (e.g., remove all vertical lines and any excess horizontal lines, center all values, etc.).

-You mention that those in the community-focused condition were more politically conservative (M = 2.99, SD = 1.46) compared to those in the self-focused condition, M = 2.57, SD = 1.53, t(299) = -2.44, p = .015, 95% CI: [-0.76, -0.08]. Do you think this could have impacted the lack of effect that restriction condition had on reactance, given that political orientation significantly impacted reactance? It is possible I missed where this was discussed or ruled out in the analyses. But if not, please make mention of this either in your analyses or briefly in your discussion.

-Why is gender listed in Fig. 3 but not Fig. 2?

-The disproportionate distribution of gender was not mentioned in your limitations section. I recommend mentioning this alongside the statement about ethnicity.

MINOR NOTES

- Throughout the introduction, the United States is sometimes written as U.S. and other times is written as US. Please keep consistent throughout.

-For figures 1a and 1b: in the text, the a and b are lowercase but, in the figures, they are uppercase (A; B). Please keep consistent.

-In some places, you round to three decimals, and in other places, you round to two. For example, the Cronbach’s alpha for State Reactance was .896, 95% CI [.867, .917]. I recommend keeping it to two decimals throughout.

-Under the “State Reactance” subheading, please change the following sentence from “The scale consisted of 11items assessing experience of reactance” to “The scale consisted of 11 items assessing experience of reactance” (insert space between 11 and items).

-You included gender in your Fig. 3 model, but it is not named in the caption for Fig. 3. Please include alongside mention of the other variables.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

10 Nov 2021

We thank the anonymous Reviewer for the thorough and helpful reviews they provided to the original manuscript. We have endeavored to respond to each of the comments raised in the revised manuscript and in this response letter. The specific ways in which we have responded to each of the comments is detailed below with our responses provided in italicized font.

Reviewers' comments:

6. Review Comments to the Author

Reviewer #2: The authors sufficiently responded to the comments from the Editor and Reviewer. I appreciate the amount of detail and effort that went into each of the responses. In particular, I appreciate the inclusion of gender in the analyses. I also appreciate the inclusion of subheadings in the introduction, as this makes for an easier read. MAJOR NOTES

Following up on the original comment from Reviewer 1, the replicability link still does not work, showing that the depository is empty.

We are able to view and download the data and associated analytic script when we click the replicability link, and we are unsure why the depository is empty for both you and the Editor. Since receiving the revision request, we have deleted and reuploaded the files, and had multiple lab members check and they were able to view and download the files. This should be remedied and allow for a complete review.

Where did the item for “Adherence to Social Distancing” come from? Is this pulled from another questionnaire, or was it developed by the authors themselves for the purpose of this study? Please state either way.

We created this item for the purposes of the study and have clarified this in the “Adherence to Social Distancing” subsection (p. 12).

Was the wording of the 6 items from Conger and colleagues’ economic hardship measures adapted to include “since the pandemic began”? If so, please state that you adapted/changed this measure to meet the needs of the current study.

We have added the following sentence to specify that the wording was changed for these measures in order to better capture financial stress in association with COVID-19, “The prompt “Since the pandemic began” was added to items to orient participants to rating financial stress experienced in the wake of the pandemic” (p. 12).

Under “Psychological Sense of Community”, you provide an example item for the two items measuring personal identification, but you do not provide an example item for the two items measuring connection. While not a huge deal, I recommend also including an example item of connection for your readers.

We appreciate the recommendation and have added the following example item of connection: “Two items measured participants’ connection to those impacted by COVID-19 (e.g., “I feel strong ties to the community affected by COVID-19”) and two additional items measured participants’ personal identification with those impacted by COVID-19 (e.g., “I identify with the community affected by COVID-19”)” (p. 13-14).

You don’t mention income as a covariate in your introduction, but you mention it in your method section, and even use it as an exclusion criterion if not answered. Perhaps briefly mention it in your “Current Study” section of the introduction.

Thank you for this suggestion; we now identify pre-pandemic income as a covariate in the Current Study section (p. 9).

Table 1 is very helpful, and I appreciate its inclusion in the manuscript. However, it does not follow APA formatting guidelines. Additionally, many of the values do not line up, such that some are flush left, and some are centered, making it harder to read. Please reformat this table (e.g., remove all vertical lines and any excess horizontal lines, center all values, etc.).

Table 1 has been reformatted to meet APA 7 and PLOS One guidelines for Tables. All values are centered, except the first column containing variables names which left-centered for ease of reading associated with long text entries (e.g., names, etc.). Additionally, all unnecessary horizontal and vertical lines have been removed.

You mention that those in the community-focused condition were more politically conservative (M = 2.99, SD = 1.46) compared to those in the self-focused condition, M = 2.57, SD = 1.53, t(299) = -2.44, p = .015, 95% CI: [-0.76, -0.08]. Do you think this could have impacted the lack of effect that restriction condition had on reactance, given that political orientation significantly impacted reactance? It is possible I missed where this was discussed or ruled out in the analyses. But if not, please make mention of this either in your analyses or briefly in your discussion.

We understand the Reviewer’s concerns and have added an Exploratory Post Hoc Analysis subsection to the Results (pp. 20-21) to address the possibility of a restriction condition x political orientation interaction impacting results (p. 20-21). Of note, including this interaction resulted in poor model fit, χ^2(30) = 274.31, p < .001, robust CFI = .74, robust TLI =.61, robust RMSEA = .19 (90% CI: [.17, .21), and the restriction condition x political orientation interaction was nonsignificant, b = -0.01, p = .85. Moreover, the main effect of restriction condition remained nonsignificant, b = .39, p = .09, while the significant restriction condition x financial stress interaction remained significant, b = -0.29, p = .03. Finally, we have added the following sentence to the Limitations section of the Discussion to acknowledge the possibility of a more politically conservative community-focused group may have influenced the restriction condition main effect: “Despite random assignment to restriction condition, the disproportionate representation of more politically conservative participants within the community-focused group may have elevated reactance within this condition and reduced the magnitude of the predicted difference between restriction conditions; future research should explore this possibility further” (p. 26). Our hope is that these additions to the manuscript better address the potential impact of political orientation on our results.

Why is gender listed in Fig. 3 but not Fig. 2?

The disproportionate distribution of gender was not mentioned in your limitations section. I recommend mentioning this alongside the statement about ethnicity.

We have added a paragraph to the Strengths and Limitations subsection that discusses the disproportionate distribution of gender as a limitation and that contextualizes our findings within the gender and COVID-19 literature (p. 26).

MINOR NOTES

1. Throughout the introduction, the United States is sometimes written as U.S. and other times is written as US. Please keep consistent throughout.

We have corrected the error and the United States should now be written as U.S. throughout the manuscript.

2. For figures 1a and 1b: in the text, the a and b are lowercase but, in the figures, they are uppercase (A; B). Please keep consistent.

We have corrected the error and have capitalized A and B within the text to maintain consistency with the figure.

3. In some places, you round to three decimals, and in other places, you round to two. For example, the Cronbach’s alpha for State Reactance was .896, 95% CI [.867, .917]. I recommend keeping it to two decimals throughout.

We have corrected the error and have all decimals rounded to two decimal places, except for p-values.

4. Under the “State Reactance” subheading, please change the following sentence from “The scale consisted of 11items assessing experience of reactance” to “The scale consisted of 11 items assessing experience of reactance” (insert space between 11 and items).

We have corrected the typological error and have inserted a space between “11” and “items” for this sentence.

5. You included gender in your Fig. 3 model, but it is not named in the caption for Fig. 3. Please include alongside mention of the other variables.

We have corrected our error and have listed gender in the figure caption.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

17 Nov 2021

We’re all in this together: Focus on community attenuates effects of pandemic-related financial hardship on reactance to COVID-19 public health regulations

PONE-D-21-19560R2

Dear Dr. Knapp,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Neha John-Henderson

Academic Editor

PLOS ONE

15 Dec 2021

PONE-D-21-19560R2

We’re all in this together: Focus on community attenuates effects of pandemic-related financial hardship on reactance to COVID-19 public health regulations

Dear Dr. Knapp:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Neha John-Henderson

Academic Editor

PLOS ONE