1. Introduction
The Coronavirus pandemic 2019 (COVID-19) spread rapidly all over the world and has posed serious public health challenges worldwide [
1,
2]. The public has been experiencing not only physical health problems but also psychological crisis [
3,
4]. As the number of infectious cases and mortality rate rapidly increased, fear of COVID-19 leads to individual-level mental distress such as anxiety, depression, post-traumatic stress disorder, and suicide, but also some psychosocial problems, including stigmatization, discrimination, disruption of community interactions [
5,
6]. Therefore, screening public fears related to COVID-19 and then taking further intervention will be beneficial for both preventing psychosocial problems and carrying out governmental public health interventions.
The Fear of COVID-19 Scale (FCV-19S) was developed in an Iranian context in 2020 and shown to have strong reliability and validity scale for assessing fears related to the coronavirus [
7]. The final version of FCV-19S was a single-dimensional scale with 7 items and was shown to be significantly correlated with depression and anxiety, making it helpful for identifying these comorbid disorders [
7,
8,
9]. Subsequently, the FCV-19S has been translated into eighteen different languages [
8]. Most of these studies showed that it is a unidimensional scale. However, studies proposed a two-factor structure, such as in Israeli sample [
10], Ecuadorian sample [
11], Chinese population sample [
12] and Russian adolescents [
13]. These inconsistent results also show that it is an unstable factor structure of the FCV-19 Scale [
14]. Meantime, most studies were small or middle-aged samples. These included work in Iranian (N = 717, mean age: 31) [
7], Italian (N = 249, mean age: 34) [
14], Saudi (N = 639, mean age: 35) populations [
15]. These discrepancies may be due to differences in sample characteristics, cultural backgrounds, or experiences of the COVID-19 epidemic, including different, countries, ethnic groups, epidemic control situations, and so on. Therefore, the FCV-19 scale’s psychometric properties should be further studied across different cultures or vulnerable samples, especially in the elderly, adolescents, and clinical samples [
16,
17].
In China, approximately 80,000 individuals have been diagnosed with COVID-19, with over 4600 officially recorded deaths (Chinese National Health Commission 2020). The massive infectious public health event has put enormous pressure on the Chinese government, health care providers, and the public [
18]. Level 1 public health response was activated in 31 Chinese provinces [
19]. There are 33.66 million college students nationwide, including 8.83 million inter-provincial students. The continuous spread of the epidemic, strict isolation measures and delays in starting schools, colleges, and universities across the country were expected to influence the mental health of college students [
20,
21]. Existing studies have found that university students were more vulnerable to the harmful effects of media information overload of the COVID-19 outbreak, including panic, anxiety, and depression [
22,
23,
24]. Incidences of anxiety and depression among Chinese university students were up to 40–50% during the COVID-19 epidemic [
1,
25,
26,
27]. One of the responsibilities of universities is to protect the physical and mental health of students and prevent the possible consequences of the spread of the epidemic [
28]. During the COVID-19 epidemic, the National Health Commission of China issued a number of measures to reduce the spread of the virus, such as lockdowns, quarantine, and online teaching, and implemented emergency psychological crisis intervention for the public [
29]. Meantime, the Ministry of Education of China has issued guidelines on mental health services in universities, including screening, monitoring the mental health status, and increasing the number of full-time and part-time psychological counselors [
30].
Although the epidemic has been well under control in China, sporadic outbreaks have still occurred with most of the infected cases arising from southwest border cities or villages. Therefore, the present study will cross-culturally adapt and validate the FVC-19S in university students, which will contribute to mental health care in universities.
4. Discussion
This study assessed the cultural adaptation of the C-FCV-19S and found it to be psychometrically valid for Chinese university students and to help quickly screen the fear of COVID-19. This study had three principal findings. (1) The finalized C-FCV-19S had two dimensions and 8 items. (2) The C- FCV-19S had good reliability and validity, with an optimal cutoff point of 17.5. (3) The C-FCV-19S score was a positive association with anxiety and depressive symptoms.
Ahorsu et al. developed the FCV-19s as a timely self-evaluation measurement meant to assess fear of COVID-19 during the pandemic. The final version of the FCV-19s was a unidimensional scale with 7-item [
7]. Other research in different countries validated the scale and had inconsistent results. Studies in Italian and Arabic general populations [
14,
15] and Spanish university students also demonstrated that the 7-item version had a unidimensional structure [
45]. However, another study, such as in Japanese adolescents [
46] and university students of Russia [
13] showed a bi-factor scale. It is worth noting that two studies for the Chinese sample so far were inconsistent. Specifically, one study of Chinese students, including schools and universities reported a single-dimensional structure scale [
9], while another study of the general Chinese population found a two-dimensional structure scale [
12]. However, our study found that the 8-item C-FCV-19s was a bi-factor model. One reason for these differences may be sample characteristics and size. We collected a large sample of university students (N = 2334), whose average age was 19 years old (range: 18–26), while other studies of the Chinese population used teenagers and adults (N = 1700; mean age: 18; range: 10–57) [
12]. In the present study, the two deleted items were “I am afraid of losing my life because of Coronavirus-19” and “When watching news and stories about Coronavirus-19 on social media, I become nervous or anxious”. These descriptions may not fit well with the current situation of the epidemic in China, as the COVID-19 epidemic has overall been well-controlled and Chinese people tend to believe in the government’s ability to deal with COVID-19 [
47]. More importantly, this difference may come from the specificity of college students as a group. It has been illustrated that young people with higher levels of moral disengagement are less likely to engage in preventive behaviors in public crisis situations such as the COVID-19 pandemic [
48,
49]. In addition, 7 and 8 items of the FCV-19S belong to threat perception. It seems that in young people higher levels of perceived threats concerning personal health or the health of loved ones did not correspond to a greater probability to adopt preventive behaviors [
50]. One possible explanation is that adolescents are at a much lower risk of contracting the most severe symptoms of COVID-19 [
51] and therefore they consider the probability of being infected with serious harm to be very low. In other words, threat perception does not influence the adoption of healthy behaviors by adolescents in the face of the COVID-19 epidemic [
50]. Taken together, we have good reason to speculate that two items are excluded from the CFA analysis.
Our study supported the idea of satisfactory reliability and validity of the two-dimension structure of the C-FCV-19S. Firstly, EPA found that two factors of the C-FCV-19S, based on factors with eigenvalues, were greater than those of random datasets. Further, CFA suggested that the C-FVC-19S was consistent with the data, which corroborates previous research indicating that the data model fits the two-factor structure. In the SEM, the model fit GFI, TLI, and CFI indices were all obtained to the ideal value (
n > 1000), although the RMSEA values and χ
2/df were above the ideal value. Some studies have suggested that this was acceptable in large sample studies (
n > 1000) [
40,
52]. Traditionally, the accepted standard of fit indices (GFI, TLI, CFI, etc.) > 0.9 or above to be acceptable, and models with RMSEA between 0.05 and 0.08 are acceptable [
41]. However, recent research suggests that the analysis of these indices was based on MLE (Maximum Likelihood Estimate) and CLS (Generalized Least Squares), and it was recommended that SRMR should be combined with TLI, RNI, or CFI indicators to test SEM fits [
40,
53]. Meantime, if the sample is above 1000 (N > 1000), the chi-square value (χ
2) is usually so large that models with good fit are rejected. The index of χ
2/df is used when the sample is less than 1000 (N < 1000) [
41,
54]. Therefore, there is good reason to conclude that our model has the goodness of model fit to support the two-factor of the C-FCV-19S.
Unfortunately, to date, only a study reported that the cutoff scores of the Greek FCV-19S version were 16.5 or higher for distinguishing elevated fear and normal of COVID-19 [
8]. We used a ROC analysis to determine the optimal cutoff point of 17.5. Most studies of FCV-19s in different countries showed that the mean score of FCV-19s was close to 17.5, with ranged from 15.6 to 18.3 [
11,
16,
45,
55]. Among them, the mean score of Russian university students was 18 ± 4.5 [
13]. Thus, these studies illustrated that the cutoff score of our study is reasonable and credible.
This study has several limitations that should be noted. First, the representativeness of the sample was a limitation. We only surveyed university students in southwest China, where COVID-19 was most prevalent. Second, confounding factors, including students’ hometowns, family health, and disease history should be fully considered. Finally, some participants may have had depression and anxiety prior to the COVID-19 pandemic, which may have influenced their fear of COVID-19. However, we were unable to collect this information in the present cross-sectional study. Therefore, the criterion validity of the C-FVS scale in this study has yet to be verified.