INTRODUCTION
When COVID-19 started spreading in Norway in early 2020, the government, with strong support from the Directorate of Health and the Norwegian Institute of Public Health, launched the most comprehensive health education effort ever seen in Norway. Information about spread of the virus and precautions that needed to be taken were communicated continuously, in all possible media channels and with heavy appeal over several months. There were no obvious and visible disagreements among the central level communicators. From a health education research perspective it is interesting to examine the compliance with the recommendations advocated and its relationship with demographic factors. Knowledge about differences in compliance across population segments and insights into processes determining compliance are important for our ability to cope with future pandemics and health crises.
A series of data collections on COVID-19-related behaviour were carried out by the Norwegian Institute of Public Health. The present paper is based on a data collection which took place in two Norwegian counties in November and December 2020. The purpose is to examine the association between level of education and behavioural precautions to avoid transmission of the COVID-19 virus among the adult population.
Studies have generally confirmed that there is a strong and mostly consistent association between level of education and health-related behaviour. A series of reports from large-scale data collections in Norwegian counties have demonstrated that this is indeed the case. Higher levels of education are associated with healthier eating, more physical activity, and less smoking. There are two main exceptions. Sedentary behaviour (number of inactive hours per day) is more widespread among the well-educated, and so are moderate levels of alcohol intake [1].
Associations between indicators of socioeconomic status, such as level of education, and health-related behaviour are generally in line with a well-established model for how innovations spread in populations [2]. Higher status and well-educated groups tend to change first and take the roles of “pioneers”, “innovators”, and “early majority”. High education has not always been associated with healthy behaviour. Cigarette smoking, probably more researched from a diffusion of innovations perspective than any other health-related behaviour, started in highly educated population segments. But, when smoking education and control programmes emerged, well-educated groups were the first ones to quit [3].
There are a few studies which have reported on the relationship between level of education and COVID-19-related behaviour. In a study from Pakistan, Balkhi and associates found no significant differences between educational groups on a number of social distancing and hygiene outcomes. There was only one exception. Avoiding or reducing use of healthcare facilities was significantly and slightly more common among those with the highest educational level (graduates) [4]. In a study from Latvia, Šuriņa and associates did not find a significant association between level of education and a sum score based on items covering social distancing and hygiene [5]. In a German sample, Hoenig and Wenz found significant differences between groups defined by level of education in the expected direction for four out of six COVID-19-related behaviours (avoided crowded places, kept distance from others, reduced social contacts, washed hands), but differences were small [6]. In a study carried out among young adults in Switzerland, non-compliance with hygienic recommendations was found to be highest among those with the highest level of education [7]. Given the modest evidence base and potential public health relevance, further studies are needed.
The purpose of the present study is to describe dimensions of COVID-19-related behaviour and the relationship of these behaviours with level of education at a relatively early stage of the COVID-19 pandemic (November-December 2020). Associations between the dimensions of coronavirus-related behaviour and gender and age will also be described.
MATERIAL AND METHODS
SAMPLE AND DATA COLLECTION
Data were collected between the 18th of November and the 4th of December 2020 in two Norwegian counties, Oslo and Vestland. The data collection occurred at the beginning of the second infection wave in Norway. In response to the wave, the government had recently (5th of October) reintroduced stricter nation-wide infection control measures, after a period of a relatively low level of restrictions. The Norwegian Institute of Public Health was responsible for the data collection, conducted in agreement with and funded by the county authorities. Representative samples were drawn from the population registry. The Common Contact Register was used to obtain electronic contact information (e-mail/SMS) and exclude those who had not allowed their contact information to be used in for instance surveys. Individuals with no electronic contact information available were also excluded. After two reminders the response rates were 39.8% in Oslo and 37.3% in Vestland (N = 15,071).
Data collections were carried out among adults (18 years +) only and based on informed consent from study participants. Ethical approval was provided by the Regional Committee for Medical and Health Research Ethics, Northern Norway (Project 458542).
INSTRUMENTS
In this paper, 23 items on COVID-19-related precaution behaviour were used.
During the last two weeks I have washed my hands or used hand disinfection fluids…
…before preparing or eating food;
…after visiting the toilet (including changing diapers);
…when my hands are visibly dirty;
…after contact with body fluids (for instance after blowing my nose);
…after contact with animals;
…when I arrive in the workplace/kindergarten/school;
…after having touched surfaces in crowded places including the workplace/kindergarten/school;
…before putting on or after removing a face mask;
I have been coughing or sneezing in a paper handkerchief or elbow hook;
I have as much as possible avoided touching my face when among other people;
I have kept at least one meter distance to other people (other than those close to me);
I have avoided handshaking or hugging other people (other than those close to me);
I have avoided crowded places during leisure time;
I have avoided going to the supermarket;
I have been to the supermarket only at times when there are few other customers there;
I have avoided public transport;
I have avoided meeting friends during leisure time;
I have avoided visiting others or having other people visiting me;
I have avoided meeting family members that I don’t live with;
I stay at home when I feel ill (including with low levels of symptoms).
I have used a face mask when I…
…have not been able to keep at least one meter distance from others, such as when travelling by collective transport;
…have been indoors in public places (stores, malls) where it has been impossible to keep others at least at one meter distance;
I have tried to keep at least one meter distance to other people, even when I have used a face mask.
Response options were ”Yes, always”, “Mostly”, “Sometimes”, “Almost never”, and “No”. In analyses of dimensionality and for the purpose of producing meanscores, the categories were coded from 0 (zero) for “No” to 5 for Yes, always”. For the purpose of presenting distributions, three response categories, “Seldom”, “Almost never”, and “No” were collapsed into one category.
A simple question on gender only presented two response alternatives, “man” and “woman”. An age variable was grouped into the following categories: 18-29, 30-39, 40-49, 50-59, 60-69, and 70+. A question on the highest completed education presented four response alternatives: (i) Primary school, Secondary school, Folk high school, (ii) Vocational education, High school, at least 3 years (iii) College or university less than 4 years, (iv) College or university, 4 years or more.
DATA ANALYSES
Statistical data analysis was done with percentage distributions, principal components analysis (pairwise deletion of cases and oblique rotation) and general linear modelling (GLM) in SPSS version 27.
RESULTS
The proportion of women in the sample was 56.2%. The mean age was 46.7 and standard deviation 15.4 years. Percentage distributions of COVID-19-related behaviour are shown in Table 1. The proportion who “Always” take a specific precaution varies widely, from 93.3% to 1.5%.
Component loadings from the principal components analysis of the whole set of behavioural variables are shown in Table 2. For the most part, only loadings as high as 0.40 or larger are shown. When a variable had no loadings as high as 0.40, the largest loading is shown. It seems reasonable, and consistent with the eigenvalue criterion, to distinguish between five components: (i) Hygiene, (ii) Hygiene when hands are dirty, (iii) Avoiding meeting people, (iv) Keeping distance and avoiding crowded places, and (v) Wearing a face mask. One variable did not load sufficiently high on any of the factors: “I stay home when I feel ill (including with low levels of symptoms)”. Another single variable contributed to reducing the alpha value considerably when included in the “Keeping distance” scale namely: “I have avoided public transport”. Meanscores (sum scores divided by number of items) were constructed for each component. The two items that were not included in any of the meanscores were analysed as single items. Descriptives for the meanscores are shown in Table 3.
An overview of scales with Cronbach α values is shown in Table 3. For three of the scales, Hygiene, Avoiding others, and Wearing face masks, α values were adequate (0.76, 0.74, 0.72, respectively). For two scales, Hygiene when hands are dirty and Wearing a face mask, values were below recommended levels (0.58 and 0.56, respectively).
The five meanscores as well as the two single items were analysed against gender with adjustments for age group. Women scored significantly higher than men on Hygiene (Cohen’s d = 0.44), Hygiene when hands are dirty (d = 0.32), Staying home when ill (including with low levels of symptoms) (d = 0.25), Wearing a face mask (d = 0.23), Avoiding contact with others (d = 0.17), and Keeping distance from others (d = 0.16). There was no gender difference at all for Avoiding public transport (d = 0.00).
There were a few significant interaction effects between gender and age group when analysed against the five behaviour meanscores and the two single items. The pattern of change over age groups was, however, rather similar for men and women. For the sake of simplicity, we therefore report differences across age groups for men and women combined, but with adjustment for gender. Mean score increased with age group for Keeping distance from others (Cohen’s d 18-29 vs.70+ = 0.39), Hygiene (Cohen’s d 18-29 vs. 60-69 = 0.25), Staying home when ill (even with low levels of symptoms) (Cohen’s d 18-29 vs.70+ = 0.12), and Hygiene when hands are dirty (Cohen’s d 18-29 vs. 60-69 = 0.10). For the other behavioural outcomes, the pattern of change across age groups was less monotonous. Avoiding contact with others was lowest among the youngest (18-29) and highest among the middle-aged (40-49) (Cohen’s d 18-29 vs. 40-49 = 0.10). The mean score on Wearing a face mask was highest among the youngest ones and lowest in the age group 50-59 (Cohen’s d 18-29 vs. 50-59 = 0.19). The mean score on Avoiding public transport did not differ much over the age span 30 years or older, but was particularly low among the youngest (Cohen’s d 18-29 vs. 60-69 = 0.46).
Mean scores on all behavioural outcome variables by education and gender and with adjustment for age are shown in Table 4 and Figure 1. Mean score on Hygiene decreases with increasing level of education for both genders, and when the group with the lowest level education is compared with the group with the highest level of education the decrease is the same for both genders (Cohen’s d lowest vs. highest level of education = –0.31). A decrease in mean score with increasing level of education is also found for Avoiding public transport (Cohen’s d lowest vs. highest level of education = –0.15). Again, the estimates were the same for both genders. A decrease in mean score with increasing level of education is also found for Avoiding meeting others and Keeping distance, but only among women (Cohen’s d lowest vs. highest level of education = –0.17 and –0.09, respectively). No significant association is found between Hygiene when hands are dirty and level of education. A positive association is found between level of education and Wearing a face mask for both women and men (Cohen’s d lowest vs. highest level of education = 0.30 and 0.41, respectively) and between level of education and Staying home when ill among men only (Cohen’s d lowest vs. highest level of education = 0.21).
DISCUSSION
Women tended to report higher levels of protection against virus transmission compared to men. This is consistent with findings from other studies [8-10]. As explained in a paper by Ferrin: “No matter to which sphere of life they refer, most studies conclude that men are more risk-taking than women either because risk attitudes are attributes of masculine or feminine psychology, or because they are culturally and stereotypically learnt” [11].
Differences were rather small across age groups. For some outcomes there was an increase in compliance on outcome variables with the largest and most consistent patterns for Keeping distance from others, Hygiene (until age 60-69), and Staying at home if ill (women only). Since being infected with the COVID-19 virus increased risk of hospitalization and a fatal outcome more strongly with older age, the increasing means on behavioural precaution outcomes in this study may reflect an increasing level of fear of being infected over age.
In six out of fourteen analyses defined by gender and outcome variables, there was a significant, negative association between level of education and means on precautions to avoid transmission of the COVID-19 virus. In five analyses there was no significant association. In only three analyses, positive associations were observed. The strongest associations were with Hygiene (d = –0.31 for both genders) and Wearing a face mask (d = 0.30 among women and 0.41 among men). All other significant associations were in the range of d = 0.09 to d = 0.21, in other words small effect sizes.
The many negative and near-zero associations were unexpected. Most studies in high income countries show health-compromising behaviour other than COVID-19-related behaviour to be more widespread among the less well educated. Some support for positive associations between level of education and behavioural precautions to avoid spread of the COVID-19 virus was found in one study from Germany [6]. Our findings are, however, consistent with results from three previous studies of COVID-19-related behaviour [4, 5, 7].
There are features of the Norwegian COVID-19 campaign and the Norwegian context which may explain why less well-educated population groups to such a large extent, and in some behaviours more than highly educated groups, complied with recommendations from the authorities. First of all, the main governmental communicators, which included the Prime minister, the Minister of Health, the Director of Health, and experts from the Norwegian Institute of Public Health, all have high credibility. Vis a vis the public there was a high level of consent. Potential disagreements among the various actors were not easy to see from the perspective of the general audience. Studies have confirmed a high level of trust in authorities among Norwegians [12]. Within a context of trust in authorities and under an extremely intensive and consistent flow of information and persuasive messages, one possible explanation is that doubt and resistance may be easier to mobilize among well-educated groups. This may explain why behavioural precautions to reduce the risk of virus transmission on some points have reached higher levels among less well-educated population segments.
The higher the level of education, the higher is the proportion of people who use face masks. This is a deviation from the pattern found for most other outcomes, namely that there is no or a negative association between level of education and behavioural outcomes. So why is this deviant pattern found? Studies have shown that there are a number of problems associated with use of face masks more than with other protective behaviours such as hygiene, avoiding meeting people, and keeping distance. In a report from the World Health Organization it is maintained that use of face masks is uncomfortable and may be accompanied by physical discomfort [13]. From a study from Saudi Arabia in 2021 it was reported that use of face masks may lead to problems with breathing, pain, they are unpleasant to wear, difficult to wear for those who use glasses, and lead to skin irritation [14]. In another study it was concluded that face masks should not be used over longer periods of time due to breathing problems and skin irritation [15]. Use of face masks may be particularly problematic for people with coronary heart health problems, chronic respiratory disorders, or diabetes, and for elderly people [16]. One possible explanation of the positive association between level of education and use of face masks is that those with higher education are more strongly motivated to cope with the unpleasantness associated with use of face masks. Having low levels of education could also be associated with experiencing daily life contexts which makes it more demanding to use face masks, for instance during work.
METHODOLOGICAL CONSIDERATIONS
The main limitation of the present study is related to possible selection bias. Response rates for both counties were slightly under 40%. And non-response was higher for less well-educated groups. A recent study of the association between response willingness (responding after first contact versus responding after one or two reminders) did not, however, indicate any noticeable level of selection [17]. To what extent this also holds true when comparing those who participate with those who do not participate remains uncertain. This may be possible to find out only through a study which, on key outcome variables, compares those who participate in the data collections with a group which is recruited among non-participants.
Strengths of the present study includes large samples, use of well-tested instruments for data collection, and use of adequate statistical techniques for the analysis of data.
CONCLUSIONS
Medium strength and positive associations between level of education and behavioural precautions to avoid transmission of the COVID-19 virus were expected. These expectations were not confirmed. For most outcome measures in this study, the associations were either negative or close to zero. Several aspects of the “Norwegian COVID-19 campaign” made it distinctly different from most media-based health education efforts which have taken place in Norway. Communication took place continuously for months across all media. Messages were strong, senders had a high level of credibility, and the high level of trust in authorities in Norway may have contributed to acceptance and compliance in most groups. Positive associations were, however found on two points. The higher the level of education was, the higher was the score on using face masks, and the higher was the score on staying at home if ill (men only). Wearing face masks is inconvenient, unpleasant, and associated with skin problems. Perhaps a high level of education is advantageous for mobilizing sufficient motivation to overcome these obstacles. Due to contextual factors, for instance related to job situations, wearing face masks may also be more challenging for less well-educated groups.
As expected, women were more compliant than men, and for at least some indicators, compliance increased with age.
ACKNOWLEDGEMENTS
We wish to thank the study participants in Vestland and Oslo counties for time and efforts spent on filling in the questionnaire.
FUNDING
Data collections were funded and data administration provided by the Norwegian Institute of Public Health.
ETHICS APPROVALS AND CONSENT TO PARTICIPATE
Data collections were carried out among adults (18 years +) only and based on informed consent from study participants. Ethical approval was provided by the Regional Committee for Medical and Health Research Ethics for Norway South-East (Application number 503277). Data Protection Impact Assessment (DPIA) has been carried out by the Norwegian Institute of Public Health (No. 23/00375-1).
DISCLOSURE
The authors report no conflict of interest.
References
1. Knapstad M, Skogen JC, Leino T, et al. The public health survey in Viken. Procedures and selected findings. Norwegian language report. Oslo: Norwegian Institute of Public Health 2022. Available from: https://www.fhi.no/div/helseundersokelser/fylkeshelseundersokelser/ (accessed: 17 October 2022).
2.
Kaminski J. Diffusion of innovations theory. Theory in Nursing Informatics Column. Canadian Journal of Nursing Informatics 2011; 6(2). Available from: https://cjni.net/journal/?p=1444 (accessed: 15 October 2022).
3.
Ferrence R. Using diffusion theory in health promotion: the case of tobacco. Can J Public Health 1996; 87 (Suppl 2): S24-S27.
4.
Balkhi F, Nazir A, Zehra A, Riaz R. Psychological and behavioral response to the coronavirus (COVID-19) pandemic. Cureus 2020; 12(5): e7923.
5.
Šuriņa S, Martinsone K, Perepjolkina V, et al. Factors related to COVID-19 preventive behaviors: a structural equation model. Front Psychol 2021; 12: 676521.
6.
Hoenig K, Wenz SE. Education, health behavior, and working conditions during the pandemic: evidence from a German sample. Eur Soc 2021; 23 (Suppl 1): S275-S288.
7.
Nivette A, Ribeaud D, Murray A, et al. Non-compliance with COVID-19-related public health measures among young adults in Switzerland: insights from a longitudinal cohort study. Soc Sci Med 2021; 268: 113370.
8.
Galasso V, Pons, V, Profeta P, Foucault M. Gender differences in COVID-19 attitudes and behvior: panel evidence from eight countries. PNAS 2020; 177(44): 27285-27291.
9.
Chuang Y, Liu JGE. Who wears a mask? Econ Bull 2020; 40(4): 2619-2627.
10.
Clark C, Davila A, Regis M, Kraus S. Predictors of COVID-19 voluntary compliance behaviors: an international investigation. Global Transitions 2020; 2: 76-82.
11.
Ferrin M. Reassessing gender differences in COVID-19 risk perception and behavior. Soc Sci Q 2022; 103(1): 31-41.
12.
OECD. Government at a Glance 2021. Available from: https://dx.doi.org/10.1787/1c258f55-en (accessed: 4 September 2022).
13.
World Health Organization. Mask use in the context of COVID-19: interim guidance, 1 December 2020. Available from: https://apps.who.int/iris/handle/10665/337199 (accessed: 19 October 2022).
14.
Al Naam YA, Elsafi SH, Alkharraz ZS, et al. Community practice of using face masks for the prevention of COVID-19 in Saudi Arabia. PLoS One 2021; 16(2): e0247313.
15.
Adanur S, Jayswal A. Filtration mechanisms and manufacturing methods of face masks: an overview. J Ind Text 2020; 51(3): 3683S-3717S.
16.
Ahmad MDF, Wahab S, Ali Ahmad F, et al. A novel perspective approach to explore pros and cons of face mask in prevention the spread of SARS-CoV-2 and other pathogens. Saudi Pharm J 2021; 29(2): 121-133.
17.
Clarsen B, Skogen JC, Nilsen TS, Aarø LE. Revisiting the continuum of resistance theory in the digital age: a comparison of early and delayed respondents to the Norwegian Counties Public Health Survey. BMC Public Health 2021; 21(1): 730.
This is an Open Access journal, all articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0). License (http://creativecommons.org/licenses/by-nc-sa/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material, provided the original work is properly cited and states its license.
