Age, self-rated health, sex, education, income, children in the household, agreeableness, extraversion, neuroticism, perceived exposure risk show no relation to adherence to COVID-19 rules; conscientiousness does

Bogg, T., & Milad, E. (2020). Demographic, personality, and social cognition correlates of coronavirus guideline adherence in a U.S. sample. Health Psychology, 39(12), 1026-1036. http://dx.doi.org/10.1037/hea0000891

Abstract

Objective: The present study examined patterns and psychosocial correlates of coronavirus guideline adherence in a U.S. sample (N = 500) during the initial 15-day period advocated by the White House Coronavirus Task Force.

Method: Descriptive and correlational analyses were used to examine the frequency of past 7-day adherence to each of 10 guidelines, as well as overall adherence. Guided by a disposition-belief-motivation model of health behavior, path analyses tested associations of personality traits and demographic factors to overall adherence via perceived norms, perceived control, attitudes, and self-efficacy related to guideline adherence, as well as perceived exposure risk and perceived health consequence if exposed.

Results: Adherence ranged from 94.4% reporting always avoiding eating/drinking inside bars/restaurants/food courts to 13.6% reporting always avoiding touching one’s face. Modeling showed total associations with overall adherence for greater conscientiousness (β = .191, p < .001), openness (β = .098, p < .05), perceptions of social endorsement (β = .202, p < .001), positive attitudes (β = .105, p < .05), self-efficacy (β = .234, p < .001), and the presence versus absence or uncertainty of a shelter-in-place order (β = .102, p < .01). Age, self-rated health, sex, education, income, children in the household, agreeableness, extraversion, neuroticism, perceived exposure risk, and perceived health consequence showed null-to-negligible associations with overall adherence. 

Conclusions: The results clarify adherence frequency, highlight characteristics associated with greater adherence, and suggest the need to strengthen the social contract between government and citizenry by clearly communicating adherence benefits, costs, and timelines.

KEYWORDS: COVID-19, personality, conscientiousness, social cognition, guideline adherence

Discussion

The purpose of the present research was to investigate patterns and psychosocial correlates of adherence to the White House Coronavirus Task Force guidelines for slowing the spread of the novel coronavirus (SARS-CoV-2) using a U.S. sample. As assessed during the week following the release of the guidelines, the results showed there to be generally high, but not perfect, frequency of following most of the 10 guidelines, especially for avoiding social gatherings in groups of more than 10 people, avoiding eating or drinking inside bars, restaurants, or food courts, and avoiding visiting nursing homes or retirement or long-term care facilities. It also is notable that nontrivial minorities of participants indicated less frequent adherence to all of the guidelines as well, especially avoiding touching of the face, coughing or sneezing into an elbow, disinfecting frequently used items, washing hands for 20 s or more, avoiding being closer than six feet to other people, and avoiding social visits. These patterns reveal the variations to guideline adherence that may further contribute to the unwitting spread of SARS-CoV-2, as well as morbidity and mortality due to COVID-19.

The results of the path modeling show some of these variations can be explained by individual differences in personality traits, beliefs about guideline adherence, and, to a lesser extent, perceptions of current health. Specifically, in line with trait-consistent temperamental process models of behavior, conscientiousness was directly associated with greater past seven-day frequency of overall guideline adherence by virtue of the general tendencies to be reliable (vs. careless). Consistent with an instrumental disposition-belief-motivation perspective, open individuals were more likely to follow the guidelines by virtue of more positive attitudes associated with following the guidelines. Also consistent with an instrumental disposition-belief-motivation perspective, agreeable individuals were more likely to follow the guidelines by virtue of greater endorsement of norms and attitudes associated with following the guidelines. However, the total association between agreeableness and guideline adherence was not statistically significant. Moreover, the total associations between extraversion and neuroticism and guideline adherence also were not statistically significant.

Consistent with Social Cognitive Theory, individuals who were more confident in overcoming obstacles to following the guidelines—entreaties for social company or not feeling like it—were more likely to follow the guidelines. Consistent with the Theory of Planned Behavior, individuals who perceived others as supportive or encouraging of following the guidelines were more likely to follow the guidelines, as were individuals who held more positive views of the guidelines—as being wise or useful. In contrast, as components of the Health Belief Model, individuals who perceived greater risk of exposure and/or greater perceived health consequence were not more likely to follow the guidelines. Moreover, the results did not show consistent effects or differences for age, sex, education, income, and the presence/absence of children in the household on guideline adherence.

The findings for conscientiousness are consistent with a large body of research demonstrating the health relevance of this personality trait (e.g., Bogg & Roberts, 2013). Moreover, guideline adherence is a prototypical exemplar of conscientiousness—following socially prescribed norms and delaying gratification (Roberts, Jackson, Fayard, & Edmonds, 2009). The small total effect observed for openness is more novel, but does add to a growing body of research demonstrating the health relevance of this personality trait (e.g., Bogg & Vo, 2014; Graham et al., 2017). Similarly, the findings for perceived norms, attitudes, and self-efficacy are consistent with decades of theorizing and research using Social Cognitive Theory and the Theory of Planned Behavior. Indeed, from the vantage point of the study of individual differences in health-related behaviors, guideline adherence, despite its unprecedented status, is associated with many of the same tendencies and beliefs as other behaviors. However, with the spread of morbidity and mortality throughout the population at stake, the implications for these associations are much more acute and severe.

Implications

Given that members of the White House Coronavirus Task Force have stated they expect a resurgence of SARS-CoV-2 during the fall of 2020 and that a safe and reliable vaccine is not likely to be available until the winter of 2021, the present findings have implications for the ongoing and future use of national-level guidelines and state-, county-, and city-level emergency orders to slow the spread of the virus. To be clear, these implications pertain more to the focused and consistent implementation of existing public health approaches, rather than to wholesale changes or shifts in strategy.

One of the lessons learned from the person–situation debate within personality psychology is that consistent relations between personality traits and behaviors should not be expected in “powerful” and “clearly normatively scripted situations” (Kenrick & Funder, 1988, p. 31). It is when traits are provided with sufficient situational affordances for variable expression that covariation with behavior should be expected. As can be inferred from the results of the present study, situational flexibility was observed to the extent conscientiousness and social cognitions were found to be directly associated with guideline adherence. While these findings are validating from a construct perspective, they also show how individual differences can affect public health measures and guidance.

In the U.S, the prevailing ethical premise of public health policy during a pandemic is the use of evidence-based measures that do not unduly restrict individual liberties or harm well-being (Gostin, Friedman, & Wetter, 2020). In instances where more extreme measures, such as stay-at-home orders, are deemed necessary based on available evidence, then the affected population must be assured that basic needs (e.g., medical care, schooling, housing, income) will be provided for by the government and that such measures and their associated penalties have clear sunset provisions. In such a way, an unambiguous social contract can be established—one with both positive and negative contingencies associated with complying with the measures. As a prerequisite consideration to a social contract, the results of the present work suggest all those affected must be fully informed and/or reminded as to whether they are subject to more onerous measures, such as shelter-in-place or stay-at-home orders.

Several approaches could be used to strengthen perceptions of the binding nature of such a social contract. Early, consistent, and visible messaging regarding the nature and scope of the threats associated with transmission and infection would be required. This would entail careful coordination between public health and political leaders at all levels of government in order to frame the guidelines as necessary and legal emergency measures, rather than advisements for consideration. Coordinated messaging regarding the measures would likely help alter any (mis)perceptions that individual rights and liberties are absolute, that there is arbitrary local/regional variation in the utility or importance of such measures, and that public officials might appear to ignore, minimize, or repudiate the measures.

Establishing and maintaining a clear social contract is consistent with the goals of emergency public health measures (i.e., introducing and sustaining new norms for behaviors while mitigating collateral harms to well-being through the use of emergency measures). Under an effective social contract for such measures, the influences of individual differences would likely remain, but could be reduced. In principle, the terms of the social contract should serve as the primary influences of guideline adherence. In such a context, a primary task of the political–public-health apparatus would be establishing and strengthening perceptions of a social contract. To the extent there is a perception of a stronger set of contingencies for guideline adherence, then there should be a reduction in the influence of the individual characteristics associated with adherence. As noted, such a perspective is consistent with principles from the fields of personality and social psychology, which hold that more powerful situations tend to attenuate the influence of individual difference factors on behavior.

Clear articulation and sustained communication of the following could serve to strengthen perceptions of a social contract for adherence behaviors:

1. The benefits of adherence—the offsetting means by which the collateral effects of emergency measures on individual and institutional well-being would be mitigated.
2. The costs of nonadherence—aside from risks of infection and illness, the precise consequences for violations of the emergency measures and assurances that individual and institutional violators should expect them to be fair and certain.
3. The limited timeframe for adherence—the necessity of emergency measures will be continually reevaluated and emergency orders for such measures will be rescinded at the earliest appropriate opportunity.

The above recommendations are not intended to be exhaustive, but illustrate example means by which the perceived influence of the situational constraints surrounding social distancing and hygienic measures can be strengthened via explicit social contract. This will remain a concern, given the subsequent implementation of additional measures (e.g., masks), as well as fluctuations in restrictions based on changes in local rates of coronavirus infections. Clarifying the existence, structure, and contingencies of such a contract could help reduce gaps in adherence associated with lower conscientiousness and weaker beliefs about adherence behaviors.

Limitations

Although the approach of the present work provides some clarity and insights into the patterns and correlates of guideline adherence in the U.S. during the initial 15-day period of guideline implementation, the results do not come without limitations. First, the representativeness of the sample by age, sex, and race was inherently limited. The sampling strategy available from Prolific did not allow for further stratification by income, education, region, and so forth, or many other characteristics and features of the population used to strengthen claims of representativeness. Second, because the approach of the study emphasized assessment during the initial 15-day period, obtaining approval from the relevant institutional review board was prioritized. This resulted in an approach that avoided survey questions that could potentially be personally identifiable (e.g., ZIP codes), violate HIPAA or other relevant privacy regulations (e.g., symptomatic/diagnosed family members), or otherwise pose a risk greater than everyday life (e.g., reporting maladaptive coping behaviors). This approach was effective in obtaining exempt status in a timely manner, but also resulted in a more limited assessment of candidate psychosocial correlates of guideline adherence. Third, the precision of the guideline items and scale was limited by its retrospective framing and self-report format. Moreover, although the scale demonstrated adequate rudimentary psychometric properties, a more sophisticated probing of its structure is warranted. Fourth, intention (planning) to follow the guidelines, while an integral component of the Theory of Planned Behavior, was excluded due to the cross-sectional design of the study, which precluded the appropriate temporal ordering of intention prior to behavior. Finally, prospective and longitudinal designs would allow for tests of temporally predictive effects to guideline adherence, as well as COVID-19 symptoms and diagnoses, rather than relying on the tests of associations reported in the present work.