Table 4.

Effect on study outcome: Model comparisons

Mo-del	predictor	ß	SE	95% CI			p	AIC	BIC	χ² (∆ df)	∆ p
1	-	-	-	-		-		3011.2	3026.2
2	GPA	0.321	0.029	0.263	0.378	***		2898.4	2918.3
Model 1 vs. Model 2										114.842( 1 )	***
3	TMS	0.251	0.041	0.172	0.331	***		2863.4	2888.2
Model 2 vs. Model 3										37.049( 1 )	***
4	VT	0.325	0.125	0.081	0.570	**		2858.6	2888.4
Model 3 vs. Model 4										6.762( 1 )	**
5	Age	0.001	0.034	-0.066	0.068	0.977		2860.6	2895.4
Model 4 vs. Model 5										0.000( 1)	0.985
6	Sex	-0.214	0.059	-0.330	-0.098	***		2847.6	2882.4
Model 4 vs. Model 6										12.998 (1)	***
7	VT*Age	0.090	0.089	-0.084	0.264	0.310		2877.2	2921.8
Model 6 vs. Model 7										0.855 (2)	0.652

Note. Random-Intercept Models with predictors as fixed effects, DV = study outcome parameter; N = 1063; GPA = German “Abitur”; VT = completed relevant vocational training; Model 1 = Medical school as varying intercept; Model 2 = Model 1 + GPA as fixed effect; Model 3 = Model 2 + TMS as fixed effect; Model 4 = Model 3 + VT as fixed effect; Model 5 = Model 4 + Age as fixed effect; Model 6 = Model 4 + Sex as fixed effect; Model 7 = Model 6 + Age*VT interaction as fixed effect; AIC = Akaike’s Information Criterion; BIC = Schwarz’s Bayesian Criterion; ∆ p = Significance of model difference; * p < .05, ** p < .01 *** p < .001