Validation of the Canadian French version of the fear of COVID‐19 scale in the general population of Quebec

Abstract Background The purpose of this study was to develop a Canadian French translation of the fear of COVID‐19 scale (FCV‐19S) and assess its psychometric characteristics. Methods A forward and backtranslation process was conducted for the Canadian French version of the FCV‐19S. The guidance of the ISPOR task force for translation and cultural adaptation was followed and cognitive debriefing interviews were conducted with six citizens. The final proofread Canadian French FCV‐19S was then administered to a large sample of citizens from the province of Quebec in Canada through an online survey. A quota sampling was conducted in 2020. Respondents from the survey also completed the Clinical Outcomes in Routine Evaluation (CORE)‐6D and the Sense of Coherence (SOC‐3) questionnaires. Several psychometric tests were performed to investigate the reliability (internal consistency) and validity of the Canadian French FCV‐19S, including construct validity, concurrent validity, and Rasch analysis. Results The translation process was conducted without any major difficulties. The cognitive debriefing interviews led to no change in the reconciled translation. The survey collected answers from 3428 citizens. Results indicated that the factor structure of the Canadian French FCV‐19S is a unidimensional factor fitting well with the data. The scale showed adequate reliability (Cronbach's alpha of .903) and concurrent validity, as indicated by significantly negative correlation with CORE‐6D (r = –.410) and SOC‐3 (r = –.233). The Canadian French FCV‐19S properties tested using Rasch analysis was also very satisfactory. Conclusions The results of the present study indicated that the Canadian French version of FCV‐19S is a unidimensional tool with robust psychometric properties in the adult's population of all ages residing in the province of Quebec, Canada.


BACKGROUND
In April 2020, the 2019 coronavirus pandemic , according to the World Health Organization, has become a global health emergency affecting 180 countries and territories (WHO, 2020), and generating multiple socioeconomic and psychological consequences in many countries, including Canada (Roy et al., 2020;Tandon, 2020;Xiang et al., 2020). In addition to freezing all educational and business activities for months, the global pandemic, as in other pandemics, has brought fear, anxiety, and worries among the population as major psychological consequences (Roy et al., 2020;Tandon, 2020;Xiang et al., 2020).
The uncertainty surrounding coronavirus including its clinical symptoms, treatment, and vaccination is the hardest thing to handle. People still don't know exactly how they'll be impacted, how long this will last, or how bad things might get. And that makes it all too easy to catastrophize and spiral out into overwhelming dread and panic (Ahorsu et al., 2020).
According to Arora et al. (2020), COVID-19 affects all aspects of life, and the risk factors are more diverse and numerous than in other pandemics. Therefore, fear and its harmful psychosocial consequences, as the literature points out, can go beyond illness or death (Yoon, 2020;Arora et al., 2020). And this by contracting the virus and inducing other fears such as infecting others, development of denial or phobia for infected people (Pappas et al., 2009;Falagas & Kiriaze, 2006), neuroticism in participants who have experienced a loss of a loved one (Murphy & Moret-Tatay, 2021), economic hardship (Yoon, 2020;Arora et al., 2020), the exacerbation of the infectious disease itself (Ahorsu et al., 2020;Pappas et al., 2009), and the labeling of infected people perceived to be the origin of the disease (Pappas et al., 2009;Falagas & Kiriaze, 2006).<COMP: Please set reference citations as per the journal style, that is, in alphabetical order.> Attitudes are also of interest in understanding the effects of COVID-19 outbreak. Sense of belonging to the country of reference has been shown to be related to and directly predict fear (Murphy & Moret-Tatay, 2021). Additionally, the protective measures such as lockdowns and quarantines associated with the pandemic COVID-19 have been shown to have major effects on stress and fear (Lu et al., 2020;CDC, 2020).
As the fear of encountering individuals who may have been infected has been reported in the context of COVID-19 (CDC, 2020a,b;Lin et al., 2020), a new psychometric assessment tool assessing an individual's fear of COVID-19 was recently developed by Ahorsu et al. (2020), that is, the fear of COVID-19 Scale (FCV-19S), a short and valid robust assessment scale (Lin et al., 2021), across different populations. This scale has been used in countries such as Iran (Alyami et al., 2021), France , Bangladesh (Sakib et al., 2020), Italy (Soraci et al., 2020), Turkey (Satici et al., 2021), Russia and Belarus (Reznik et al., 2021), Israel (Tzur Bitan et al., 2020, Peru (Huarcaya-Victoria et al., 2022), and Paraguay (Barrios et al., 2021). In addition, in the province of Quebec, the rising fear of COVID-19 among nursing staff was measured using the validated French-Canadian (French-CA) and English-Canadian (English-CA) versions of the fear of COVID-19 scale (FCV-19S) (Gélinas et al., 2021). Even though the FCV-19S was validated in a large sample of nursing staff, additional validation testing with other multidisciplinary members is required to substantiate its validity in the Quebec health care workforce (Gélinas et al., 2021

Transcultural adaptation and validity
The FCV-19S, originally developed in Farsi (Ahorsu et al., 2020) and then in English (Winter et al., 2020), was translated from the English version by experienced translators into French. It is a seven-item scale that assesses the fear of COVID-19. The seven items (e.g., "I am most afraid of coronavirus-19") are rated on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree) with scores ranging from 7 to 35. The higher the score, the greater the fear of COVID-19. In this study, we followed the principles indicated by the ISPOR task force for translation and cultural adaptation (Wild et al., 2005

Participants and procedure
Participants were invited to take part in the study via an online survey that was conducted by the survey company Dynata inc. Quota sampling was performed in 2020 among the general population according to age, gender, and educational level. Inclusion criteria for volunteers were being (i) at least 18 years old, (ii) French-speaking citizens, (iii) residents of the province of Quebec, Canada, and (iv) agreeing to participate in the study. All participants completed the survey anonymously and gave their informed online consent. All procedures conducted were approved by our local ethics committee review board at the CIUSSS de l'Est de l'île de Montréal.

Measures
Questions concerning sociodemographic aspects of the participants (e.g., age, gender, educational level, marital status, occupation, income, body mass index [BMI], urban or rural, country of origin, residency status) were included in the online survey. This was done along the use of self-reported questions on health and life satisfaction, and risk acceptance on a 11-level scale.
The CORE-6D scale (Mavranezouli et al., 2013) is a 6-item health descriptive system consisting of a 5-item unidimensional emotional component (e.g., I never feel terribly alone and isolated) and a physical item (I am never troubled by aches, pains, or other physical problems).
Each item has three levels. The unidimensional emotional component of CORE-6D, combined with the physical item, creates a 2-dimensional scale, tapping emotional, and physical symptoms in people with com-mon mental disorders. As many items in the CORE-6D relate to fear and anxiety, it was considered appropriate to use this instrument to test the concurrent validity of the FCV-19S.
The Sense of Coherence (SOC-3) scale (Schumann et al., 2003) is a concept that reflects the ability to cope with stress and is at the core of the salutogenesis theory. It consists of at least three dimensions: the comprehensibility, the manageability, and the meaningfulness components (e.g., Do you usually see a solution to problems and difficulties that other people find hopeless?). The 3-point SOC-3 is the simplified 3-item version, which was used in our study to measure the sense of coherence. The sum score of the SOC-3 ranged from 0 to 6, and higher values indicated a higher sense of Coherence. The

Data analysis
The statistics were analyzed using SPSS V.26 (for descriptive statistics, PCA, concurrent validity) and WINSTEP (for Rasch analysis). Descriptive statistics were used to describe the study participants' characteristics.
The psychometric properties were first examined as follows: descriptive statistics of the FCV-19S items (i.e., means and confidence interval of the main items); item-intercorrelations (Spearman's ρ), and corrected item-total correlations (Spearman's ρ between the respective item and the total score without the respective item). An acceptable value for the corrected item to-total correlation should be >.4 (Wang et al., 2007).
For reliability of the FCV-19S, an internal consistency calculation was carried out using Cronbach's alpha which, with a value greater than 0.7 indicates satisfactory internal consistency and good test reliability (Nunnally & Bernstein, 1994;Lin et al., 2018a). Also, a Cronbach's alpha was calculated by excluding one item at a time. If values increased by more than 2%, the item was considered potentially problematic.
Principal component analysis for construct validity was conducted using SPSS V.26 to measure the factor structure of the Canadian French FCV-19S. The aim of this analysis was to examine the dimensionality of the FCV-19S (Bagozzi & Yi, 1988;Lin et al., 2019). In this study, eigenvalues 0.8 or higher were used to create a matrix with varimax rotation and obtain one component. The weight attributed to each item helped identify items with a low contribution to the variance.
Items with a weight above 0.40 were considered good for a component explaining more than 20% of the total variability. For a component explaining less than 20% of the variability, items were considered good if their contribution was above 0.70.
For concurrent validity, we examined the association of the FCV-19S with the Sense of Coherence scale 3-item and the CORE-6D scale using the Pearson's correlation coefficient (r). Correlation coefficients less than .25 were considered as poor correlation; between .25 and .5 as moderate; between .5 and .75 as good; and higher than .75 as excellent correlation.
The following properties of the FCV-19S were examined using the Rasch model: item statistics, item and person separation reliability, item and person separation index, and differential item functioning (DIF). Rasch analysis is a statistical technique traditionally used for binary data, but some polytomous generalizations can be used for interval data (Lin & Pakpour, 2017). Explaining at least 50% of the variance in the Rasch dimension, and an eigenvalue of less than 2.0 on first contrast, provides evidence of unidimensionality (Linacre, 2012).

Translation and transcultural validation results
In the reconciliation process of the two forward translations from English to Canadian French, 10 modifications of words or sentences were made for a better sounding in Canadian French. The recon-

Linguistic validation
The linguistic validation was carried out with 6 citizens-3 men and 3 women-aged between 24 and 67 years. Mean age was 42 years (SD ± 16.88). Five out of six were employed (83.3%) and one was retired. Four have high educational background with a graduate university diploma and two have secondary education level. All respondents were com-fortable with the questionnaire, did not have any difficulty understanding the instructions and the response options made lots of sense for them. The meaning was as expected, and they did not have any suggestions to help improving the questionnaire.

Psychometric validation
A total of 3547 citizens consented to participate in the study and 3428 were included for analysis. A total of 119 were excluded for the following reasons: missing data, aged less than 18, and nonresidents of Quebec. Data collection occurred from July to November 2020 and mean time to complete was around 15 min.   The present study also analyzed the distribution of the seven FCV-19S items. Two items (3 and 6), as shown in Table 2, were asymmetrically distributed. As for asymmetry and Kurtosis, two items, which do not fall in the range of -1 and +1, were distributed in a nonnormal way (Tabachnick & Fidell, 2007 all items (Table 3)  The psychometric testing has also shown that the FCV-19S has a high internal consistency with a Cronbach's alpha of .903 for the sevenitem scale, which could not be improved by removing any items.  Table 5), which is also satisfactory.
Additional analysis also indicates that age was negatively correlated with the FCV-19S but was not significant (r = -.016, p = .170). However, results have shown a significant positive correlation between gender and the pattern of FCV-19S (r = .085, p < .01).

DISCUSSION
The present study detailed the process of translating the fear of Psychometric analyses showed that the CF-FCV-19S has a good internal reliability and consistency when assessed by Cronbach's alpha (.903). This is in line with previous internal consistency findings from other countries (Iversen et al., 2021;Ahorsu et al., 2020;Alyami et al., 2021;Tzur Bitan et al., 2020;Reznik et al., 2021;Sakib et al., 2020;Satici et al., 2021;Soraci et al., 2020;Tsipropoulou et al., 2020;Winter et al., 2020, Huarcaya-Victoria et al., 2022Pang et al., 2020) reporting a Cronbach's alpha between .82 and .89. As indicated by the results (Table 3), each individual corrected item correlated >.70 with the total score, indicating satisfactory internal consistency (Everitt & Skrondal, 2010). to previous studies, it has been stated that the health of individuals can be affected by negative psychological states such as anxiety during periods of infectious epidemics (Duncan et al., 2009;Pappas et al., 2009;Ropeik, 2004). Therefore, the CF-FCV-19S could help the population better understand the emotional factors associated with pandemics and subsequently help reduce anxiety and stress (CDC, 2020a,b).

STRENGTHS AND LIMITATIONS
Psychometric tests of the CF-FCV-19S demonstrated that the CF-FCV-19S is a reliable and valid tool to assess the severity of fear of COVID-19 in Quebec adults of all ages. The study was also conducted in a quite representative and large sample of the general population in Quebec. However, the present study has certain limitations. First, since the assessment of the concept of fear is subjectively carried out; however, the social desirability factors that may influence participants' responses must be considered. To mitigate this bias, an online survey without human interaction can be performed. Second, although selfreport data are a valuable information resource for constructs, selfreport bias must nonetheless be taken into account (Althubaiti, 2016).
Third, in surveys, there is always a possibility of self-selection bias (Bethlehem, 2010), which may indicate an overrepresentation of specific subgroups within the general population. In addition, the sample in the current study did not include adolescents and children. Finally, participants were from the general population residing in the province of Quebec and no formal diagnosis on mood disorders was obtained (e.g., anxiety). Therefore, the sensitivity and specificity of the scale cannot be examined. Future studies should thus evaluate if individuals with underlying medical conditions associated with a higher risk of death from COVID-19 (e.g., diabetes, hypertension, coronary heart disease, preexisting respiratory conditions) may experience increased levels of COVID-19 fear.

CONCLUSION
The translation and transcultural validation of the CF-FCV-19S led to an easy-to-read and -understand questionnaire. In addition, the psychometric testing of the CF-FCV-19S demonstrated that the instrument is a unidimensional reliable and valid tool for assessing the severity of fear of COVID-19 among adults of all ages residing in the province of Quebec.

DATA AVAILABILITY STATEMENT
Data will be available upon reasonable request to the corresponding author.