**3. Results**

Both tests for a common method bias showed that no such bias was present. The variance explained by a single-factor model was only 14.72%, much lower than the threshold of 50% (Harman's one-factor test). Furthermore, all the correlation coefficients between our variables were between 0.01 and 0.65 (absolute value), that is smaller than the threshold value of 0.90 (correlation matrix test).
