Luppo Kuillman

Moral reasoning explained by personality traits 63 3 fact that participation was voluntary; and c) that they could end their participation in the study at any time. Respondents were also informed that their answers were completely anonymous and would not be used for any purpose other than the study. Furthermore, the letter clearly addressed the expected average completion time (45 minutes) for filling out the questionnaires. This study was performed in accordance with the tenets of the Declaration of Helsinki (General Assembly of the World Medical Association 2014). Only the first author (LK) had access to the encrypted data. The questionnaires were filled out using Qualtrics software (version January-December 2016). All questions were forced-choice, which produced no information bias due to missing values in scale construction and statistical analysis. Data analysis Prior to analysis, all continuous variables (except the standardized N2-index) were transformed into approximately normal distributed indicators by the two-step transformation process (Templeton & Burney 2016). Bivariate analysis For categorical data, we used the chi-square test (Fisher’s exact tests for 2 x 2 contingency tables) and the difference between proportions test (Newcombe & Altman 2000). For continuous variables, we used student t-test for independent samples. Multivariate analysis Structural equation modelling (SEM) only allows testing of unidirectional effects among multiple continuous dependent and independent variables. Therefore, SEM was more appropriate for our study than standard multiple regression techniques because it allows simultaneous assessment of the strength and direction of the interrelationships. To test the theoretical model against the observed data, we applied SEM. The structural model concerned the direct and indirect effects of Stability and Plasticity on moral reasoning. Both the direct and indirect effects (through moral disengagement) were estimated in a path analysis. As our sample size ( N = 155) was rather small, we decided to perform a bootstrapped SEM by replicating the sample twice. As