Question

I've lost count how many times I posted this question, at least 4 or 5. Can someone please help me out

1. In her paper on measuring the returns to high school sports, Betsey Stevenson investigates the causal implications of expansion in female sports participation caused by Title IX. Compliance with Title IX can be characterized as requiring a school to raise its female athletic participation rate to near equality with its male athletic participation rate. The paper is here: http://www.nber.org/papers/w15728.pdf She first looks at cross sectional relationship between high school sports, and labor market outcome:s sing 1979 National Longitudinal Survey of Youth (NLSY). (You should not have to read the paper to answer this question). Consider the following table from the paper: Table 1 Effects of High School Participation in Extra Curricular Activities on Educational Attainment and Log Wages Independent Variable Female Male Panel A: Dependent Variable: Years of Education (OLS) Athletics 632 436 471 380 (117) 09.098 119 12 097) 096) 358 .095) -,362 .101) 441 984 862 429 Non-vocational Clubs 462 .104) -.456 Vocational Clubs 103) 499 Adjusted R-squared 117 260 431 165 297 .485 Panel B: Dependent Variable: Log of Wages (OLS) 072 Athletics 106 076 190 142 076 073 039)039 038) 039034 034034 033) 018 Non-vocational Clubs 010 033 .061 036) 277 .039) 032 Vocational Clubs .040) 211 210 142 202 275 Adjusted R-squared .092 140 Controls Demographics Family characteristics School characteristics Ability/achievement Source: Author's calculations based on data from National Longitudinal Survey of Youth, 1979 (NLSY79). Education was measured in 1994 when respondents were 29-37 years old. Standard errors robust to heteroskedasticity in parentheses.) Sample is restricted to those who completed at least 10th grade. Participation in extra-curricular activities was asked in 1984. Athletics is an indicator variable for an individual having participated in high school sports Non-vocational clubs include student government, newspaper, yearbook, and other, primarily hobby, clubs. The National Honor Society is not included among the clubs, but is included as a control for academic ability/achievement. Demographic controls include a saturated set of dummy variables for age, race, urban status at age 14, and state of residence at age 14. Family characteristics include parents' education (measured as the highest grade completed by either parent), whether respondent lived with both parents in high school, number of siblings, family poverty status in 1978, and whether the household had a newspaper subscription or a library card. School characteristics include the percentage of eachers with a masters degree, the percentage of students who are disadvantaged, the dropout rate, and the attendance rate. Ability/achievement controls include AFQT score, membership in the National Honor Society, and self-reported measures of self-worth and failure Consider the second regression (column 2) What is the dependent variable in Panel B? a. b. What is the hypothesis being tested in Panel B? (note that this regression includes other controls too) Which variable captures the hypothesis? c. d. What type of a variable is this (hint: the footnote b under the table)? What is its coefficient? e. f. What does the coefficient mean? Is it significant at the 5% level? g. h. What does this mean? Would you say that there is good evidence for an effect of high school athletics on log wages? Explain i. Bonus question: Why do you think this estimate on sports participation might be an over-estimate (i.e. this estimate might be bigger than the true estimate)? Hint: Think about omitted variable bias

Answer 1

A) The dependent variable is Wages

B) The hypotheses are:

Ho: B1 = 0 (The coefficient is equal to zero)

Ha: B1 =/ 0

C) The variable is the Athletics variable whether the individual has participated in high school sports

D) It is a categorical variable or a dummy variable with 2 levels.

E)The coefficient is 0.106

F) It means that the effect of having an individual participated in high school sports increases the expected wages by a multiple of e^(0.106) or 1.11

G) Let's find the t-statistic which is Coefficient/Standard Error = 0.106/0.039 = 2.72

df = 6283 - 2 = 6281

The critical value is 1.96. As the t-statistic calculated is more than 1.96, we can reject the null hypothesis.

Therefore, the coefficient is significant at 5% level.

H) This means that the Athletics variable is a significant predictor of wages.

I) Yes, as the coefficient of athletics in the regression equation is significantly greater than zero at significance level of 0.05, we can say that there is a good evidence for an effect of high school athletics on log wages.