Participants
This study used a national sample by implementing an online survey based on the South Korean population censusstandard in 2018, such as age, gender, and residential area. Initially, 2440 individuals entered the survey and 988 subjects who did not meet the criterion of the present study or did not complete the survey were excluded, indicating completion rate of 59.5%. Lastly, 18 participants were excluded due to careless responses by screening partially random or inattentive data. Thus, a total of 1434 participants were used for the final analysis. Among the total sample, 731 (51%) were men and 703 (49%) were women. The mean age of the participants, who ranged from 19 to 84 years of age, was 44.34 (SD = 13.93). All participants met the following eligibility criteria: They were able to read and write Korean proficiently, were able to provide informed consent, and were aged 19 years or older. Table 1 provides sociodemographic information as well as COVID-19-related information of the total sample in more detail.
Procedure
The survey was conducted via an Internet survey company between February 19 and March 3, 2021. The number of confirmed COVID-19 cases and deaths in South Korea during the survey period were 91,236 and 1,612, respectively. During this period, government regulations that mandated social distancing, banned private gatherings of more than five people, and ensured that restaurants and bars were closed after 10 pm were in place in the nation’s capital area. In addition, the AstraZeneca and Pfizer vaccination campaigns had begun in Hong Kong, Nepal, the United States, and Japan for high-risk groups (e.g., people with chronic diseases, medical staff, and the old and infirm). The participants were assured that their data would remain confidential and anonymous, and their informed consent for participating in this study was subsequently taken. The survey took approximately 20–30 minutes to complete, and a certain amount of online credit points were provided to the participants as compensation. The questionnaire consisted of two sections. The first section asked about participants’ sociodemographic information and COVID-19-related experiences, whereas the second section included the PCL-5 and different sociological and psychological scales. The survey company is certified by ISO 9001, indicating that it meets the most recognized quality management system standards. To ensure the security of the survey, the company used a firewall (WAF) and DigiCert security service. Moreover, all survey responses were collected through an encrypted secure socket layer (SSL), which enabled the authentication, encryption, and decryption of data. At the end, all the data was removed securely once the operation of the system expired. The current study was approved by the Institutional Review Board (IRB) of the university that the researchers are affiliated with and all methods were performed in accordance with the relevant guidelines and regulations.
Table 1
Sociodemographic and COVID-19-related information of the participants
Variables
|
Total Sample (N = 1434)
|
Sex
|
Men
|
731 (51.0)
|
Women
|
703 (49.0)
|
Age
|
19–29 years
|
275 (19.2)
|
30–39 years
|
267 (18.6)
|
40–49 years
|
317 (22.1)
|
50–59 years
|
319 (22.2)
|
60–69 years
|
219 (15.3)
|
> 70 years
|
37 (2.6)
|
Occupation
|
Student
|
121 (8.4)
|
Office worker
|
731 (51.0)
|
Medical practitioner
|
53 (3.6)
|
Self-employed
|
157 (10.9)
|
Housewife
|
205 (14.3)
|
Unemployed
|
121 (8.4)
|
Others
|
47 (3.3)
|
Residential Area
|
Capital area
|
695 (48.4)
|
Other metropolitan area
|
289 (20.0)
|
Medium and small sized cities
|
450 (31.6)
|
Socio-economic level
|
Upper middle class
|
56 (3.9)
|
Middle class
|
602 (42.0)
|
Lower middle class
|
776 (54.1)
|
Household type
|
One-person household
|
228 (15.9)
|
Group household
|
1206 (84.1)
|
COVID-19-related experiences
|
Similar symptoms
|
48 (3.3)
|
Cohort isolation
|
2 (0.1)
|
Quarantine
|
35 (2.4)
|
Infected
|
2 (0.1)
|
No symptoms
|
1347 (93.9)
|
COVID-19-related experiences of family and acquaintances
|
Symptoms similar to COVID-19
|
75 (5.2)
|
Cohort isolation
|
7 (0.5)
|
Quarantine
|
157 (10.9)
|
Infected
|
68 (4.7)
|
No symptoms
|
1127 (78.6)
|
Measures
PCL-5 To measure the level of the participants’ PTSD symptoms, we used the Korean version (Park et al., 2020) of the Posttraumatic Stress Disorder Checklist (PCL) (Weathers, Litz, Herman, Huska, & Keane, 1993), which later applied the diagnostic criteria of the DSM-5 (Blevins, Davis, Witte, & Domino, 2015). The PCL-5 has a total of 20 items, and the sub-factors are re-experiencing (five items; e.g., painful and unwanted memories about the stressful experience repeatedly come to mind), avoidance (two items; e.g., avoiding memories, thoughts, or emotions related to the stressful experience), negative alterations in cognition and mood (seven items; e.g., difficulty remembering important parts of the stressful experience), and hyperarousal (six items; e.g., nervousness, anger, externalizing behavior, or explosive/aggressive behavior). Responses are provided using a 5-point Likert scale that ranges from “not at all” (0 points) to “very much” (4 points). Higher scores indicate more severe PTSD symptoms. According to Lee, Gu, Kwon, and Lee (2020), the Cronbach’s alpha coefficients were .92, .91, .93, .93 for re-experiencing, avoidance, negative alterations in cognitions and mood, and hyperarousal, respectively. In this study, the PCL-5 showed good internal consistency with Cronbach’s alpha coefficients of .93, .88, .90, and .91, and a Composite Reliability (CR) of .94, .88, .90, and .91 for re-experiencing, avoidance, negative alterations in cognitions and mood, and hyperarousal, respectively.
K-PC-PTSD In this study, we utilized the Korean version of the PC-PTSD-5 (K-PTSD-5) scale that was developed by Yum (2017) to screen for PTSD symptoms. Originally, the PC-PTSD-5 scale was developed by Prins et al. (2003) and revised by Prins et al. (2016). The K-PTSD-5 consists of five items as a single factor, with items scored dichotomously as either “yes” (1 point) or “no” (0 point). Higher scores indicate a higher risk of symptoms, and the cut-off point for high-level PTSD symptoms is estimated to be 3. The Cronbach’s alpha coefficient at the time of the scale’s development was .73 (Prins et al., 2003), and Cronbach’s alpha coefficient in this study was .66.
Somatization The revised Patient Health Questionnaire (PHQ-15) by Kurt, Spitzer, and Williams (2002) was used to assess the pattern and severity of physical symptoms. The PHQ-15 consists of 15 items extracted from the PHQ (Spitzer, Kroenke, & Williams, 1999). Each item is scored on a 3-point Likert scale that ranges from 0 (not bothered at all) to 2 (very distressed). The cumulative score ranges from 0 to 45, with a higher score indicating a higher level of physical symptoms. The Cronbach’s alpha coefficient of the Korean version of the PHQ-15 was .73. In this study, the Cronbach’s alpha coefficient was .87.
Depression The Center for Epidemiological Studies Depression Scale (CES-D)—a self-reporting simple screening test tool developed by the American Institute of Mental Health in 1971—was used to examine the participants’ level of depression. The scale was originally validated by Cho and Kim (1993), and the short Korean version of the CES-D-10 was standardized by Shin (2011). The scale consists of ten items, and participants were asked to answer the questions, which pertained to the symptoms of depression experienced over the past week, with either “yes” (1 point) or “no” (0 points). The cut-off point that indicates a significant level of depression was estimated to be 3. In this study, the Cronbach’s alpha coefficient was .83.
Anxiety The Generalized Anxiety Scale (GAD-7)—developed by Spitzer et al. (2006) and later validated in Korean by Seo and Park (2015)—was used to identify the anxiety level of the participants and the probable cases of generalized anxiety disorder. Seven items that asked about participants’ anxiety and worries related to the COVID-19 crisis were rated by them using a 4-point Likert scale ranging from 0 (not at all) to 3 (nearly every day). A higher total score indicates a higher severity of anxiety symptoms, with an optimal cut-off point of 5. Out of a total score of 21, 5 or more, 10 or more, and 15 or more are classified as mild, moderate, and severe anxiety symptoms, respectively (Spitzer et al., 2006). In the validation study conducted by Seo and Park (2015), the Cronbach’s alpha coefficient was .92. In this study, the Cronbach’s alpha coefficient was .93.
Posttraumatic Anger The dimensions of anger Reactions-5 (DAR-5), which was developed by Forbes et al. (2014), was used to measure the level of anger symptoms. The instrument was first translated into Korean by bilingual researchers and was later back-translated by a professor of counseling and PhD-level researchers. Any discrepancies were noted and discussed until the final version was completed. This scale has a total of five items: frequency, intensity, duration, aggression, and interference with social relations. On the original scale, participants were asked to respond while recalling their daily lives over the past 4 weeks. However, in this study, participants responded while thinking about the difficulties that they experienced in their daily lives during the COVID-19 pandemic in order to measure individual anger symptoms related to the COVID-19. The participants responded using a 5-point Likert scale ranging from 1 (none of the time) to 5 (all of the time). Higher scores reflected worse symptoms of anger. The Cronbach’s alpha coefficient for all items in the DAR-5 was .90, indicating a high level of reliability. In this study, the Cronbach’s alpha coefficient was .91.
Negative Affect The Positive and Negative Affect Schedule (PANAS) scale, which was developed and validated by Hong (2004) based on the circumstances in Korea, was used. The PANAS is a widely used checklist that reflects two subscales that contain 11 items of positive affect and 11 items of negative affect. Each item is scored on a 5-point Likert scale that ranges from 1 (not at all) to 5 (very much). As the purpose of this study was to measure the negative affect of citizens during the COVID-19 crisis, the 11 items of negative affect were extracted for use. The Cronbach’s alpha coefficient of the Korean version of the PANAS (Hong, 2004) was .90. In this study, the Cronbach’s alpha coefficient was .93.
Work Burnout We used the Maslach Burnout Inventory-General Survey (MBI-GS) that was developed by Schaufeli et al. (1996) to measure job burnout. The original MBI-GS consists of 16 items, including five items that measure exhaustion, five items that measure cynicism, and six items that measure professional efficacy. A validation study of the South Korean version (Shin, 2003) that consisted of only 15 items was referred to, and the one remaining item was translated and back-translated by PhD-level researchers. All items are scored on a seven-point scale; higher scores on exhaustion and cynicism and lower scores on professional efficacy indicate a higher level of burnout. In Shin’s study, the Cronbach’s alpha coefficients for exhaustion, cynicism, and professional efficacy were .90, .81, and .86, respectively. In this study, the Cronbach’s alpha coefficients for exhaustion, cynicism, and professional efficacy were .92, .90, and .92, respectively.
Suicidal Ideation To assess the degree of suicidal ideation, a Korean validation study (Suh et al., 2017) of the depressive symptom inventory-suicidality subscale (DSI-SS)—a subscale of the Hopelessness Depression Symptom Questionnaire (Metalsky et al., 1997)—was used. The items asked about the frequency, intensity, controllability, and content of suicidal thoughts. Each item is rated on a 4-point Likert scale (0–3 points), and the total score ranges from 0 to 12. Higher scores are indicative of a greater severity of suicidal ideation. The Cronbach’s alpha coefficient of the Korean version of the DSI-SS was .93. In this study, the Cronbach’s alpha coefficient was .95.
Data Analysis
Descriptive statistics were used to analyze the participants’ characteristics, and a normality test was subsequently conducted to determine if the data followed a normal distribution. A CFA was conducted to evaluate four potential structural models of the K-COVID-PTSD scale based on theoretical and empirical evidence of PTSD. First, the single-factor model where all items were loaded in one general factor was tested. The DSM-5 four-factor model, which included re-experiencing, avoidance, negative alterations in cognitions and mood, and hyperarousal, was tested next. We then examined the third model—a six-factor anhedonia model that consisted of re-experiencing, negative affect, anxious arousal, dysphoric arousal, avoidance, and anhedonia. The final model that was tested was a seven-factor model that was suggested by Armour et al. (2015); it included re-experiencing, avoidance, negative affect, anhedonia, externalizing behavior, anxious arousal, and dysphoric arousal. A CFA was conducted considering the maximum likelihood (ML) estimation. Model fit indices of each model were examined using the chi-square test, comparative fit index (CFI), Tucker-Lewis index (TLI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA). Additionally, the Akaike information criterion (AIC) and Bayesian information criteria (BIC) index were used to compare between different models (Kass & Raftery, 1995; Raftery, 1995). The reliability analysis was followed by a Cronbach’s alpha analysis. It presented a value of more than .7, which indicated good internal consistency. Lastly, the concurrent validity of the scale was verified via Pearson correlations between measures of PTSD, somatization, depression, anxiety, posttraumatic anger, negative affect, job burnout, and suicidal ideation, respectively. The data were statistically analyzed using SPSS 21.0 and Mplus 8.0.