Table 2. Responding Family Medicine Residency Program Directors’ Support of Preference Signaling by Supplemental Application Support.

Panel A: Distribution of responses (N=262)
I would be in favor of making a supplemental application a standard component of the recruitment process for family medicine residency applicants.	I would be in favor of family medicine residency applicants taking part in preference signaling, where applicants can designate a limited number of programs that are of particular interest in their residency search.
	Not agree*		Agree
	N (%)		N (%)
Disagree	60 (71)		86 (48)
Agree	24 (29)		92 (52)
Pearson χ2(4)=287.971; P=.000
*“Not agree” includes those who responded Neither agree nor disagree, Somewhat disagree, and Strongly disagree.
Panel B: Multivariable logistic regression results (N=241)
Variables	Adjusted odds ratio*	95% CI
Support supplemental application
Yes	3.155	1.507	6.604
No/not sure/missing (ref)	-	-	-
Community size
Less than 30,000 (ref)	-	-	-
30,000 to 74,999	0.163	0.029	0.919
75,000 to 149,000	0.151	0.030	0.770
150,000 to 499,999	0.175	0.036	0.860
500,000 to 1 million	0.187	0.032	1.100
More than 1 million	0.125	0.023	0.677
Total residents
Less than 19 (ref)	-	-	-
19-31	0.466	0.214	1.014
More than 31	0.567	0.185	1.739
Program director gender
Female (ref)	-	-	-
Male	0.358	0.170	0.755
Other/nonbinary	0.604	0.039	9.388
LR χ2(41)=52.66; Prob>χ2=0.104. Log likelihood=-123.11; Pseudo R 2=0.1762.Abbreviations: CI, confidence interval; LR, likelihood ratio. *Adjusted for type of residency program, residency program location, community size, total residents, residency program director (RPD) degree, RPD current tenure, RPD gender, RPD ethnicity, RPD race, underrepresented in medicine (URiM) faculty percentage, URiM resident percentage, and resident gender percentage. Only 241 respondents had complete demographic information to be included in the regression model.