Table 2. Estimates from Generalized Linear Mixed Models for effects of academic qualfication, empoyment in healthcare, and location on test performance.

All results based on generalized means, averaged over effect of test timing (pre-, post-). 95% CI’s for contrasts are not adjusted for multiple comparisons.

Academic Qualification
Level	Proportion answers correct	SE	95% CI
Post	0.794	0.0240	0.743, 0.837
School	0.797	0.0492	0.683, 0.877
Under	0.774	0.0287	0.712, 0.825
Contrast	Odds ratio	SE	df	95% CI
post / school	0.982	0.329	174	0.507, 1.90
post / under	1.127	0.242	174	0.737, 1.72
school / under	1.148	0.394	174	0.583, 2.26
Work in Healthcare
Level	Proportion answers correct	SE	95% CI
No	0.773	0.0212	0.729, 0.812
Yes	0.816	0.0291	0.752, 0.867
Contrast	Odds ratio	SE	df	95% CI
No / Yes	0.766	0.172	177	0.493, 1.19
Location
Level	Proportion answers correct	SE	95% CI
Dundee	0.726	0.0354	0.651, 0.790
Edinburgh	0.837	0.0261	0.779, 0.883
Hull	0.715	0.0395	0.631, 0.786
USA	0.835	0.0272	0.774, 0.882
Contrast	Odds ratio	SE	df	95% CI
Dundee / Edin	0.516	0.133	175	0.310, 0.859
Dundee / Hull	1.060	0.277	175	0.633, 1.774
Dundee / USA	0.525	0.137	175	0.313, 0.880
Edin / Hull	2.054	0.554	175	1.206, 3.499
Edin / USA	1.018	0.275	175	0.598, 1.733
Hull / USA	0.495	0.135	175	0.289, 0.849