Fig. 6

Tests for confounding factors in the models. To check for confounding factors the models were built by adding in one independent variable at a time. In model 1 (labeled a in the figure), where individual resilience is the dependent variable and mindfulness and organizational resilience are the independent variables, mindfulness might be a confounding variable. This is indicated by a change in the β-value (and standardized β-value) that is rather large. However, the large CI makes this change less worrisome. In model 2 (labeled b in the figure), where organizational resilience is the dependent variable and individual resilience and mindfulness are the independent variables a similar challenge occurred. This can indicate that the confounding might be between mindfulness and individual resilience. However, the CI is still large enough that the change in value in mindfulness is not worrisome. In the third and last model (labeled c in the figure), mindfulness still might be a confounding variable with individual resilience based on the change in its β-value when individual resilience is introduced which is not seen when organizational resilience is introduced to the model. The change in beta-value is the largest in this model, and the smaller CI makes this change more worrisome than in the other two models. The change in β-values and large CI can also, in part, be explained by the correlation between these factors and the relationship between them that has been established in previous studies. Since the evidence for confounding is not that strong and the indication of confounding is between two factors with a known correlation the choice was made to perform the statistical analysis as planned.