Table 12.

Testing the proportionality of coefficients—experienced measures (p values)

	ELSA		Gallup			HWB12
	Troubled/fatigue	Positive	Troubled	Positive	Fatigue	Troubled	Positive
ELSA positive	0.02
Gallup troubled	0.47	0.04
Gallup positive	0.20	0.77	0.88
Gallup fatigue	0.85	0.97	0.96	0.99
HWB12 troubled	0.43	0.02	0.66	0.04	0.01
HWB12 positive	0.16	0.59	0.79	0.92	0.22	0.04
HWB12 fatigue	0.19	0.33	0.82	0.67	0.09	0.19	0.89

The null hypothesis tested here is $H_0=\frac{{\mathit{\beta }}_{1,model1}}{{\mathit{\beta }}_{1,model2}}=\frac{{\mathit{\beta }}_{2,model1}}{{\mathit{\beta }}_{2,model2}},$ etc. therefore testing the proportionality of coefficients across pairs of models. The table shows p values of the test statistics corresponding to the null hypothesis for each pair of models