TABLE IV.

Effect of Public TReatment On Sign-up Decision for Students Taking One to Three Honors Classes

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Dependent variable	Dummy: student signed up for the SAT prep course
Public treatment	−0.1673*** [0.061]	−0.1486** [0.061]	−0.1465** [0.063]	−0.1467** [0.064]	0.0850* [0.047]	0.0729 [0.046]	0.0834* [0.045]	0.0887* [0.048]
Inference robustness (p-values)
Robust standard errors	.007	.017	.022	.023	.071	.115	.064	0.067
Wild bootstrap	.008	.008	.028	.032	.114	.106	.084	0.070
Permutation test	.006	.027	.030	.031	.061	.116	.069	0.046
Mean of private take-up		0.848				0.864
Includes individual covariates	No	Yes	Yes	Yes	No	Yes	Yes	Yes
Includes classroom and surveyor fixed effects	No	No	Yes	Yes	No	No	Yes	Yes
Includes honors subjects variables	No	No	No	Yes	No	No	No	Yes
Observations	184	176	176	176	159	155	155	155
Sample		Nonhonors classes				Honors classes

Notes. This table restricts the sample to students taking one, two, or three honors classes. Columns (1) to (4) restrict the sample to nonhonors classes, and columns (5) to (8) restrict to honors classes. Columns (1) and (5) present OLS regressions of a dummy variable for whether the student signed up for the SAT prep course on a public sign-up dummy, controlling for dummies on the number of honors classes taken by the student. Columns (2) and (6) add individual covariates (age and dummies for male and Hispanic). Columns (3) and (7) further add surveyor and classroom fixed effects. Columns (4) and (8) replicate columns (3) and (7) adding controls for the number of honors classes taken by subject categories (math/sciences and social sciences; the omitted category is humanities). Robust standard errors in brackets. *** p<.01, ** p<.05, * p<. 1. See Section III.D for a description of the wild bootstrap and permutation test procedures.