Participants and procedure
Facebook users aged ≥ 20 years and living in Taiwan were recruited into this study between April 10th and April 23rd 2020. A Facebook advertisement was posted, which included a headline, main text, a pop-up banner and a web-link to the research questionnaire website. The recruiting advertisement was designed to appear in the “News Feed” of Facebook, which is a streaming list of updates from the user’s connections (e.g. friends) and advertisers. A previous study indicated that News Feed advertisements are more effective in terms of recruitment metrics for research studies [12]. In order to increase its visibility, we also posted the online advertisement to the social medias, such as Line and Facebook group.
This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital (approval no. KMUHIRB-EXEMPT(I)20200011). Although participants were not given any incentive for their participation, at the end of the questionnaire we provided them with the URL links to the online COVID-19 Information Centers from the Taiwanese CDC, Kaohsiung Medical University Hospital, and the Medical College of National Cheng Kung University so they could search for useful information.
Questionnaires
Perceived social support
We estimated the levels of satisfaction with individuals perceived social support using three questions: “In the past week, did you receive satisfactory support from your 1) family, 2) friends, and 3) colleagues or classmates?” The responses were graded on a 5-point Likert scale, with scores ranging from 0 (entirely disappointed) to 4 (extremely satisfied). Higher total scores indicated more satisfaction with their level of perceived social support during the COVID-19 pandemic.
Active coping with COVID-19
In a previous study we developed 7 questions to assess the respondents’ level of active coping with the threats of COVID-19 during their daily lives [13]. These questions asked participants if they: 1) avoided going to crowded places, 2) kept good indoor ventilation, 3) cleaned or disinfected their house more often, 4) washed their hands more often, 5) wore a mask, 6) searched for information on COVID-19, and 7) avoided clinic visits or missed reservations at clinics in the past week. The responses were transformed into 0 (“no” or “yes, but not due to COVID-19”) and 1 (“yes, due to COVID-19”).
Risk perception toward COVID-19
According to Liao et al [13], we developed the following question to assess the severity of current worry towards COVID-19: “Please rate your level of current worry towards COVID-19.” The severity of current worry towards COVID-19 was rated from 1 (minimal) to 10 (extremely severe). We also developed 4 additional questions to evaluate the different categories of risk perception: 1) “If you developed flu-like symptoms tomorrow, would you be worried? Reply: 1 (not at all) to 5 (extremely)”, 2) “In the past week, have you worried about catching COVID-19? Reply: 1 (not at all) to 5 (extremely)”, 3) “How likely do you think it is that you will contract COVID-19 over the next 1 month? Reply: 1 (impossible) to 7 (guaranteed)”, and 4) “What do you think your chances are of getting COVID-19 over the next month compared with others outside your family? Reply: 1 (impossible) to 7 (guaranteed)”.
Confidence against COVID-19
Self-confidence about COVID-19 and perceived confidence in the local governments controlling the COVID-19 pandemic were assessed using the following 2 questions: 1) “How confident are you that you will overcome the threats of the COVID-19 pandemic?” and 2) “How confident are you that your city is controlling the COVID-19 pandemic?” The response was scored on a 5-point Likert scale, with scores ranging from 0 (not at all confident), 1 (not very confident), 2 (neutral), 3 (confident), and 4 (very confident). Higher scores indicate that the individual was more confident about overcoming the COVID-19 pandemic.
Statistical analysis
To examine the hypothesized multiple mediation model for the association between perceived social support and active coping with COVID-19, which was mediated by risk perception or confidence (Fig. 1), the following analyses were conducted using SPSS and AMOS version 23.0 for Windows (SPSS Inc., Chicago, IL, USA). We examined bivariate associations among the variables using Pearson's correlation coefficient (r). Then, the two steps of structural equation modeling (SEM) were used. First, confirmatory factor analysis (CFA) was used to verify the association between latent variables and their indicators in the measurement model. Each question was composed of observed variables (indicators), and latent variables, which indicated perceived social support, active coping with COVID-19, risk perception, and confidence. Factor loading was used as an index to assess the scale reliability between indicators and the corresponding latent variables in the CFA. In addition, Cronbach’s α was reported to examine the internal consistency reliability. The range was considered acceptable if Cronbach’s α was > 0.5 [14].
Latent variable path analysis with maximum likelihood parameter estimations was used to estimate the model adequacy and the direct/indirect effects of perceived social support on active coping with COVID-19 through risk perception or confidence [15]. As a multiple mediator model, both mediators were applied into the model to assess and compare the mediating effects. As there was a relatively high proportion of females in the study cohort and the fact that the Kolmogorov-Smirnov test (p < 0.001) for age was significant, indicating that there was non-normal distribution, age and gender were also included within the multiple mediators’ model as covariates to adjust for their effects on the latent variables. Gender (female, male and transgender) was transformed into two dichotomous dummy variables (male vs. female; and transgender vs. female) for the analysis. The standardized estimates (beta coefficient; β) were reported for the predictive strength explained in the model.
We used the Sobel test to examine the mediating effect [16]. Furthermore, to test the adequacy of the model, multiple indices were applied to verify the goodness of fit. For each of these fit indices, the values indicating an acceptable model fit were as follows: Goodness of Fit Index (GFI ≥ 0.9); Adjusted Goodness of Fit Index (AGFI ≥ 0.9); Root-Mean Square Error of Approximation (RMSEA < 0.08); and Standardized Root Mean Square Residual (SRMR ≤ 0.08) [17, 18].