Hierarchical regression analyses
To test our hypotheses, we conducted three separate hierarchical regression analyses based on the entry method. The baseline model (step 1) of the three hierarchical regression analyses consists of six personal characteristics (socio-demographic variables). A first hierarchical regression analysis (table 3) was conducted to assess the impact of control variables (step 1) and the direct impact of red tape (step 2) on resignation. A second hierarchical regression analysis (table 4) assesses the direct impact of PSM dimensions while taking into account control variables. A third hierarchical regression analysis (table 5) includes all the independent variables of the study to assess their respective contributions to the explanation of the variance of resignation. The interaction between red tape and PSM dimensions is captured and assessed at step 4 by entering the product of the respective PSM dimensions and red tape. According to the conventions of moderation models, independent variables were centred prior the formation of the product term to reduce multicollinearity . Hierarchical regression analysis is a highly suitable statistical method for separately measuring the impact of independent variables on work outcome. The results of the analyses were converted to standardized regression coefficients and are presented in tables 3–5. The statistical significance level was set at p < .05, as determined by a two-tailed test. As shown in table 3, the baseline model (step 1) (R2 = .094; p