Table 3.

Results of eight logistic regression models predicting whether a specialty is top three choice for Year 1 students (only factors with p < 0.01 shown)

Dependent variable	Independent variable	B	OR [95% CI]	p-value
Anaesthesia
	Glasgow [ref. Aberdeen]	−1.105	0.331 [0.185-0.592]	p < 0.001
Emergency medicine
	Dundee [ref. Aberdeen]	0.586	1.797 [1.192-2.710]	p = 0.005
	Glasgow [ref. Aberdeen]	0.684	1.983 [1.399-2.810]	p ≤ 0.001
	Other ethnicity [ref. White]	-.642	0.526 [0.355-0.780]	p = 0.001
General practice
	Dundee [ref. Aberdeen]	−0.647	0.524 [0.361-0.759]	p = 0.001
	Glasgow [ref. Aberdeen]	−0.436	0.647 [0.476-0.879]	p = 0.005
	Male gender [ref. Female]	−0.555	0.574 [0.452-0.729]	p ≤ 0.001
	Other ethnicity [ref. White]	−0.959	0.383 [0.274-0.536]	p ≤ 0.001
	Intellect not extremely important [ref. Extremely important]	0.786	2.196 [1.601-3.011]	p ≤ 0.001
	Work-life balance not extremely important [ref. Extremely important]	−0.627	0.534 [0.401-0.711]	p ≤ 0.001
Medical specialties	Work-life balance not extremely important [ref. Extremely important]	−0.759	0.468 [0.347-0.631]	p ≤ 0.001
Surgical specialties
	Dundee [ref. Aberdeen]	0.494	1.640 [1.140-2.359]	p = 0.008
	Male gender [ref. Female]	0.637	1.891 [1.493-2.394]	p ≤ 0.001
Obstetrics & gynaecology	Male gender [ref. Female]	−1.374	0.253 [0.154-0.4160	p ≤ 0.001
Paediatrics	Male gender [ref. Female]	−0.871	0.419 [0.323-0.542]	p ≤ 0.001
Diagnostics	Edinburgh [ref. Aberdeen]	−0.895	0.409 [0.204-0.817]	p = 0.010

OR = odds ratio; 95% CI = 95% confidence interval.

See Additional file 1 for on-line supplementary versions of Tables 3 and 4, showing odds ratios for all the predictors in each specialty.