Table 6.

Mediation results: PROCESS macro model 4 (N = 311; 5,000 bootstrap samples).

Path	b Direct	b Indirect	SE	95% BC CI	Mediation
Extrinsic → Supervision → Satisfaction	0.33***	0.21	0.03	[0.14, 0.30]	Partial
Intrinsic → Supervision → Satisfaction	0.27***	0.11	0.03	[0.05, 0.17]	Partial

BC CI = Bias-corrected 95% Confidence Interval. Neither CI includes zero, confirming significance of indirect effects (bootstrap CIs do not yield p-values; Hayes, 2018). ***p < 0.001 applies to direct effects only. b = unstandardised regression coefficient. Direct effects in this model reflect the controlled (partial) effect of each motivational predictor on Job Satisfaction after accounting for the supervisory mediator; when summed with the indirect effects, they recover the total-effect coefficients in Table 5 (EM: 0.33 + 0.21 = 0.54; IM: 0.27 + 0.11 = 0.38). The indirect effects and CIs are the inferentially relevant quantities for H5.