Table 7

Effects on STEM Outcomes

	Math Track (1)	Advanced Math (2)	Plan STEM (3)	STEM School (4)	No Plan (5)
Female student	−0.177	−0.038	−0.124	−0.143	0.137^*
	(0.108)	(0.106)	(0.158)	(0.107)	(0.075)
Female math teacher in Grade 7	0.100	0.060	0.031	−0.041	0.076^*
	(0.069)	(0.053)	(0.088)	(0.033)	(0.040)
Female English teacher in Grade 7	−0.019	0.196	−0.115	−0.018	0.128
	(0.118)	(0.138)	(0.168)	(0.033)	(0.087)
Female Korean teacher in Grade 7	0.096	−0.014	0.146	0.022	0.054
	(0.080)	(0.068)	(0.092)	(0.064)	(0.042)
Female student × Female math teacher in Grade 7	0.051	0.097^*	0.146^*	0.073^*	−0.065^*
	(0.060)	(0.055)	(0.080)	(0.040)	(0.034)
Female student × Female English teacher in Grade 7	−0.008	−0.203^**	−0.082	−0.006	−0.100^*
	(0.068)	(0.080)	(0.121)	(0.035)	(0.053)
Female student × Female Korean teacher in Grade 7	−0.068	0.056	−0.116	0.072	−0.096^*
	(0.084)	(0.074)	(0.096)	(0.089)	(0.056)
Observations	1,635	1,456	1,043	1,721	1,640
R²	0.229	0.248	0.280	0.252	0.267
Dep. Var. Mean	0.351	0.201	0.357	0.081	0.096
Original p-value for Female student × Female math teacher	0.396	0.077	0.068	0.075
Adj. p-value Wyoung for Female student × Female math teacher	0.419	0.294	0.294	0.294
Adj. p-value Bonf for Female student × Female math teacher	0.396	0.270	0.270	0.270
Adj. p-value Sidak for Female student × Female math teacher	0.396	0.244	0.244	0.244

↵Notes: Each column represents a separate regression, controlling for school-by-math-ability-group-by-English-ability-group-by-Korean-ability-group-in-seventh-grade fixed effects. Dependent variables are indicators for the choice of the math–science track in 11th grade, taking at least one advanced math course required for admission to a STEM postsecondary degree program, self-reporting the desire to seek a STEM postsecondary degree conditional on having decided on a major, attending a STEM-oriented high school, and having no postsecondary plans. We present the results for testing multiple hypotheses on the effects of female-student-by-female-math-teacher in Grade 7. We calculate family-wise adjusted p-values based on 1,000 bootstraps clustering standard errors at school level for Westfall and Young, Bonferroni, and Sidak correction. Standard errors in parentheses are clustered at school level. *p < 0.10, **p < 0.05, ***p < 0.01.