Spss statistics 27.0 for windows

Data were analyzed using IBM SPSS Statistics 27.0 for Windows. Descriptive statistics, and individual, social, and work-related factors were used to obtain frequency with percentage or mean with standard deviation. Differences in STS on individual, social, and work-related factors were identified for normality, and analyzed using the Chi-square, independent t-test, and ANOVA. Post-hoc analysis was conducted using the Scheffé test. Hierarchical regression analysis was conducted to identify various factors that influence participants’ STS, and before the regression analysis was conducted, the basic assumptions of the regression model (normality of error, linearity, equal variance, and independence) were satisfied, and the problem of multicollinearity were checked. In Model 1, general characteristics and clinical experience, patient nursing period, and hospital type were used as inputs, and in Model 2, emotional intelligence, social support, and nursing work environment were added to Model 1 to determine the magnitude of the influence of factors that affect STS and compared.

Statistical analyses were performed using IBM SPSS Statistics 27.0 for Windows (IBM Corporation, Armonk, NY, USA).
Assumptions of normality were tested using Shapiro-Wilk tests, and if necessary, data were transformed (reciprocal transformation). Mean values were compared using Student’s t-tests or by fitting linear mixed models. If model assumptions were not met despite data transformation, their respective nonparametric equivalents were used. All post-hoc pairwise comparisons were Bonferroni-corrected to adjust for multiple comparisons. α-errors of p < 0.05 (two-sided) were considered significant. Effect sizes were reported by the partial eta-squared ( ${\mathrm{\eta }}_p^2$ ) for linear mixed models, by Kendall’s W for Friedman’s tests, and else by the correlation coefficient r. Descriptive statistics are reported as mean ± SE for normally distributed data and else as median and IQR.

The PESA and the PISA were calculated and compared between the two versions using the Wilcoxon signed-rank test. Because the previous study showed a significant positive association between the PISA and BMI, the correlation of the PISA of each version and BMI was assessed using Spearman’s rank correlation coefficient. Then, multiple regression analysis was used to assess the correlation of the PISA of each version and BMI after adjusting for age and sex. A sample size of more than 200 was considered sufficient, since multiple regression analysis with six explanatory variables was performed with 72 subjects in a similar previous study [17 (link)]. The statistical analyses were performed using IBM SPSS Statistics 27.0 for Windows (SPSS Japan Inc., Tokyo, Japan) with a significance level of 5%.