2.3 Statistical Analysis
The quantitative data used in the study were summarized as arithmetic mean ± standard deviation and qualitative data as numbers (percent). As the four dependent variables in the study, DASS-21 anxiety, depression and stress subfields, and IES-R total scores were selected, and the related data were converted into binary categorical data according to the following criteria. A cut-off of the IES-R total score ≥33 was used to reflect moderate-to-severe impact.14 Similarly, individuals with a score of 7 and above in the depression subscale, 6 and above in the anxiety subscale, 10 or above in the stress subscale cut-off points were used to determine the moderate and above psychological influence reflected on DASS-21.8 Since the number of dependent variables is four, four different binary logistic regression models were applied to the data set. Before applying the related models, variable selection algorithms based on each dependent variable were applied to the data, and independent variables considered to have no contribution to modeling were removed from the data set. As a variable selection method, LASSO (Least absolute shrinkage and selection operator)15 logistic regression technique was applied. The goodness of fit and coefficients of the created models were evaluated by Hosmer-Lemeshow (p> 0.05) and Omnibus (p <0.05) tests, respectively. In logistic regression models, significance level for model coefficients was determined as p <0.05. In the analysis, “BKSY: Information Discovery Process Software” developed by Inonu University Faculty of Medicine Department of Biostatistics and Medical Informatics was used for the data analysis.16