A multi‐centric study on validation of the Fear Scale for COVID‐19 in five Arabic speaking countries

Background The Eight‐item Fear Scale is a unidimensional scale evaluating the perceived feelings of fear associated with the thought of the coronavirus. Aim The Arabic version of this scale did not exist; hence, this study aimed to translate and evaluate the psychometric properties of the Fear Scale in participants aged 18 years and above in five Arabic countries: Egypt, Lebanon, Libya, Saudi Arabia, and Sudan by using a cross‐sectional survey design. Method The English version of the COVID‐19 Fear Scale was translated into Arabic following the guidelines and disseminated through social media. Factorial and convergent validity and internal reliability were evaluated. Results: The total number of participants was 2783; the majority was young (41.9%) and female (60.5%). Fear scores were moderate in four countries and severe in Egypt. The scale showed good structural validity, with the items explaining up to 70% of the variance. The scale items correlated significantly with the total scores, and the Cronbach alpha was above 0.9. Conclusion The study concluded that the Arabic Fear Scale is a psychometrically robust scale that can be used to evaluate the perceived feelings of fear with the thought of the coronavirus or pandemic in general.

this point, COVID-19 had developed into a pandemic associated with substantial morbidity and mortality. By the end of 2020, 61.8 million reported cases and 1.4 million deaths had been identified worldwide (World Health Organization, 2020d). The MENA region accounted for 4.6% of the global burden of COVID-19 cases and 2.9% of the global COVID19-associated mortalities (IOM UN Migration, 2020).
To slow the transmission and to avoid associated disease, pandemic mitigation steps were laid down. These measures included complete or partial lockout, travel prohibitions, limitations on mass collection, neighborhood home quarantines, physical distance measures, personal security measures, and other non-pharmaceutical interventions. In addition, this unpredictable and increasingly spreading infectious disease has triggered widespread awareness, anxiety, and depression, all of which are normal psychological reactions to the randomly evolving situation, according to the WHO (World Health Organization, 2020c).
The prevailing situation of pandemics leading to substantial morbidity and mortality, coupled with economic impact because of the lockdown enforced in many countries, has resulted in adverse psychological outcomes in the community. Based on the literature, fear is impregnated in different sections of society in response to contracting the disease, and job losses across various sectors are causing marked impairment in daily life functioning. Daily activities or situations like meeting people, traveling, and excessive information in social media serve as the triggering factors for people to develop fear. Psychological effect in the form of fear may affect all age groups including children, as reported by some studies during the H1N1 swine-flu pandemic of 2009. Children's fear of reactions about the disease was found to be correlated with H1N1-related threat information from parents or social media (Remmerswaal & Muris, 2011). Adults with co-morbidities are susceptible to severe outcomes, with some requiring intensive care.
This, in turn, has also contributed considerably to the development of irritability, anxiety, and fear (Muris & Field, 2010). Health care professionals in India, working long hours with COVID-19 patients, were reported to have health crimes originating from the fear of contracting the virus (The Times of India, 2020).
Various scales apart from the commonly used scales like GAD-7 (Al-Rabiaah et al., 2020) to measure anxiety were also developed to evaluate the psychological impact and fear. COVID-19 stress scale was developed by Taylor et al. (2020), for the measurement of stress and anxiety, with emphasis on fear due to contamination, economic consequences, xenophobia, and traumatic stress symptoms. Another scale was developed by Schimmenti et al. (2020) with the conceptual analysis of four domains of fear, namely, fear of the body, fear of not knowing, fear of significant others, and fear of inaction. Some of the studies carried out during the early phase of the COVID-19 pandemic assessed the psychological impact using the impact of Event Scale-Revised (Zhang et al., 2014), and mental health status was assessed by the Depression, Anxiety and Stress Scale (DASS-21). The Fear of COVID-19 is another scale developed by Ahorsu et al. (2020) to evaluate fear associated with contracting the virus. This same seven-item scale was validated in many languages and lastly in the Amharic language to study its psychometric properties in the Ethiopian population (Elemo et al., 2020). Fear is a subjective response and can vary in different ethnic people; an attempt has been made to measure the fear in five Arab counties, with a scale validated and tested in the Arabic language.
This study was part of an extensive international survey aimed to evaluate the effect of the COVID-19 pandemic on health, fear, and depression. The more comprehensive study was explained elsewhere (Lok et al., 2021) and briefed here.

Current study's aim
This study aimed to translate the English version of the Fear Scale into Arabic and to evaluate its psychometric properties in a large sample of Arabic-speaking individuals from five Arabic countries (Egypt, Lebanon, Libya, Saudi Arabia, and Sudan).

Linguistic evaluation of the Fear Scale
To translate and validate the Arabic version of the Fear Scale, steps recommended by Anastasi's (Anastasi, 1985) and Abdel-Khalek (Abdel- Khalek & Snyder, 2007) were adopted as follows.

Psychometric evaluation of the Fear Scale
This study was part of the international CARE project, which aimed to identify the causes of fear, fear levels, anxiety, and depression in participants recruited from over 30 countries (Lok et al., 2021). Data collection was done through a cross-sectional self-administered questionnaire between August and December 2020.

Eligibility criteria
Eligible participants were community members aged 18 years or above, free from any psychological or physical problems affecting communication, and residents of one of the study countries: Egypt, Lebanon, Libya, Saudi Arabic, and Sudan.

Sample size calculation
In the CARE project, the sample size calculation was based on estimating the prevalence of a health issue. Taking the most conservative scenario of 50%, with a 5% margin of error in a 95% confidence interval, 385 subjects were needed from each country. To account for incomplete responses, 500 subjects were targeted per country. In our application, the evaluation of the factorial validity of the 8-item Fear Scale would require 100 subjects by the rule of thumb of 80 subjects per item (Kline, 2015). Hence, the sample size should be adequate for the psychometric evaluation.

Instruments
Data were collected using an anonymous electronic survey that

Fear Scale
The Fear Scale was adapted from the Champion Breast Cancer Fear Scale (CBCFS) (Champion et al., 2004). It is an eight-item unidimensional scale for evaluating the perceived feeling of fear associated with the thought of the coronavirus. The original scale showed that the eight items fell under the concept of fear that explained 57% of the variance in exploratory factor analysis (EFA). The Cronbach alpha of the CBCFS was 0.91 (Champion et al., 2004). Each item responded on a 5-point Likert scale, and a higher total score indicated a higher fear level. There had been no cut-off point for defining high or low fear. Prior approval for translating the Fear Scale from English to Arabic has been sought from the developer.

Setting
The study was carried out in five countries in the Middle East, including Egypt, Lebanon, Libya, Saudi Arabia, and Sudan.

Ethical considerations and data protection
After securing ethical approval from the Institutional Review Boards of the five respective countries, an electronic, anonymized selfadministered questionnaire was generated using a secured survey platform. The first cover page of the questionnaire explained that the survey was related to the COVID-19 outbreak. The Informed Consent was displayed on the front page of the survey that required the respondent to agree to a consent statement before proceeding with the survey. The respondents were assured that participation in the study was voluntary and that they could withdraw from the study anytime without any penalty, and all the information would be kept confidential, and results would be reported in aggregate form.

Recruitment process
The link generated by the electronic platform was circulated among the acquaintances of the research team in each country. Social media and mobile applications such as WhatsApp were used for disseminating the link.

Data analysis
Descriptive statistics were performed on the eight items of the scale.
To examine the factorial structure of the scale, EFA was conducted with factor loadings estimated by principal axis factoring. We adopted EFA since the structure of the Fear Scale for COVID-19 has not been empirically examined. The elbow method on the scree plot was used to determine the number of factors, and the rotation method used was promax. The KMO determined the sampling adequacy, and Bartlett's test of sphericity was used to indicate the significant corrections within the variables and the matrix. Floor and ceiling percentages were calculated to measure the sensitivity and coverage of the Fear Scale total score at each end (Gulledge et al., 2019). Convergent validity was assessed by examining the association between the Fear Scale and the PHQ-4 by using the Spearman rank correlation. This was conducted with the assumption that fear would be associated with depression and anxiety (Hosseini & Khazali, 2013). To assess the internal reliability, the item-total correlation and the Cronbach alpha were calculated. A pvalue less than .05 was considered statistically significant. Statistical Product and Service Solutions (SPSS) v. 26 were used to perform this analysis.

Sociodemographic characteristics of the participants
A total of 2783 participants (497 from Egypt; 459 from Lebanon; 654 from Libya; 627 from Saudi Arabia, and 546 from Sudan) were included in this study. The majority of the participants were aged between 18 and 24 years. This age group was the highest in all the study countries, with the highest reaching 70% in Sudan. Almost two-fifths of the sample were male participants, and the majority (n = 1650, 59.3%) were single. More than half were Bachelor degree holders across the five countries, with 38.5% being students. More than 70% had never consumed alcohol, and a similar number reported never smoking a cigarette. In terms of physical activity, almost half the sample reported never being involved in vigorous exercises. This was reflected in the mean hours of 8.51 (SD = 4.65) of sitting time and 7.84 (SD = 4.52) of screen time per day. The mean number of people living in the same household was 5.81 (SD = 3.24), with the highest being in Sudan and the lowest in Egypt. In terms of medical conditions, the highest prevalence was for irritable bowel disease (7.4%), and the lowest being for stroke with only one case in Saudi Arabia, while only 27.5% reported the following up on their medical condition.

Fear among the participants
The fear scores varied significantly across the countries (F(4) = 7.23;

DISCUSSION
This was the first study that evaluated the psychometric properties of the Arabic version of the Fear Scale for COVID-19. This was done across five Arabic-speaking countries on a sample of 2783 participants.
The Fear Scale showed good validity and reliability scores in each of the five samples, as was shown in the previous evaluations of the scale on a sample of 1390 women evaluated for having a fear of acquiring breast cancer (Champion et al., 2004). Furthermore, the level of fear was comparable to the current study population showing the perceived threat of the community from the COVID-19 and the consequences of contracting the virus.
The Fear Scale is exhibited as a unidimensional scale for assessing the fear level of COVID-19. It was demonstrated by a clear one-factor structure irrespective of the whole sample or in each of the five Arabic speaking countries. The variance explained by a single factor was at least 62% which is higher than the reported 48% in the original English scale for assessing the fear level in breast cancer patients (Champion et al., 2004). Moreover, all items were strongly associated with the total score in the range of 0.66 to 0.90 which was again generally higher than the range of 0.46 to 0.78 in the original English version, despite the latter was corrected for overlap. In addition, the internal consistency is high with the range of 0.93 to 0.95, which is slightly higher than the 0.91 for the original English version. Hence, a single construct underlying the Fear Scale is affirmative.
The Fear Scale also demonstrated satisfactory convergent validity by a moderate association with the PHQ-4 that collectively assesses anxiety and depression. In our context, fear is specific to the threat of COVID-19, and both anxiety and depression are generalized reactions to this threat (Champion et al., 2004). Therefore, fear was hypothesized to be associated with anxiety and depression, despite they assess distinct constructs. On the other hand, the low floor and ceiling percentages show that Fear Scale possesses reasonable sensitivity to discriminate the fear level over its plausible range.
According to the Fear Scale, fear scores ranging between 16 and 23 are considered moderate fear scores (Champion et al., 2004). Therefore, all but one Arabic country exhibited reasonable fear of the COVID-19 while Egypt reached the severe score, followed by Lebanon, which edged extreme fear. Furthermore, scores differed significantly between Egypt and Saudi, Egypt and Sudan, and Egypt and Libya. These differences could be accounted for by other causative factors to fears in different countries (Al-ghzawi et al., 2014). Some of these causative factors could be the changes in authority and the poverty in Egypt and Lebanon, which are demonstrated in exaggeration among these countries compared to the other study countries (Abdelhalim, 2018;Dunne, 2020 results. Future studies should address the causes of the fear and nail down the triggers specific to every country at the study time. Additionally, future studies should dwell on the clinical practices and health care and mental health care availabilities in each country to provide recommendations specific to each setting. Another limitation is the availability of other Fear Scales that have been previously validated in the Arabic world. However, none of these scales specifically addressed fear in the COVID-19 pandemic.
Additionally, this short scale of eight items makes it easy for participants to complete compared to the long Fear Scales used in the Arab region, including up to 120 items. Another limitation lies in the sample of the study, with the majority being younger ages. This should be addressed in future studies with different age groups. Additionally, future development of the Fear Scale should aim to identify a cut-off score indicating fear or no fear.

CONCLUSION
Assessing the impact of the COVID-19 pandemic on fear and health among Arab countries is critical to prevent the psychological outcomes of any pandemic diseases by following effective measures. The current study's findings indicate high levels of fear among the study participants, especially in Egypt and Lebanon. The Fear Scale has shown good psychometric properties when evaluated in five Arabic-speaking countries and can be used in future studies. The future direction uses the scale among different age groups to verify the effect of fear on healthy middle-aged and older adults.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/brb3.2375

DATA AVAILABILITY STATEMENT
Data sharing agreement: All data will be made available to facilitate research reproducibility while protecting patient privacy and confidentiality.