Table 3.

Weight estimates based upon repeated-measures ANOVA model.

Variable	Level	Weight	95% confidence interval	p
Matching	Matchinga	3.74	(3.21, 4.26)	<0.001
	SOAP match, same specialtya	2.34	(1.91, 2.78)	<0.001
	SOAP match, different specialtya	1.77	(1.41, 2.13)	<0.001
Specialty	Competitiveness Highb	0.26	(0.11, 0.41)	0.002
Institution	Top 10 institutionc	0.05	(−0.07, 0.18)	0.406

^a Compared to unmatched

^b Compared to medium/low competitiveness

^c Compared to institutions not in top 10 of NIH funding