Assignment 7 Economics 31 Fall 1999

Hypothesis Testing with Regression Coefficients

Reading: Mirer Chtr. 11,12,13 Beals Chtr. 9,10,11

Hypothesis Testing with Regression Coefficients

1. Write out the steps for doing a hypothesis test, for calculating R2, and
define adjusted R2 Why are significance tests crucial to regression
analysis? The rest of this assignment will require the use of the
statistical package STATA and the data file cps22an (see Ass2 for aid in
using STATA).

2. a) Using the data in the file estimate the linear relationship between
the hourly wage, and the years of experience as an independent variable.

b) Draw a diagram of the sampling distribution of the coefficient on
experience, explain the meaning of the standard deviation, and show the mean.

c) Test the hypothesis that the coefficient, EX is = to 0. Write out the
null and alternative hypothesis, the significance level, and show all
intermediate steps.

d) Construct a confidence interval for EX, and explain the meaning of this
interval.

3. a) Using the data in the file estimate the linear relationship between
hourly wage, EX, and ED.

b) Draw a sampling distribution of the coefficients on experience, explain
the meaning of the standard deviation, and show the mean.

c) Test the hypothesis that the coefficient on EX is insignificant (ie
insignificantly different from 0). Write out the null and alternative
hypothesis, the significance level, and show all intermediate steps,

d) Construct a confidence interval for EX, and explain the meaning of this
interval.

e) Compare the R2 values from 1 and 2. Explain the meaning of R2
in the simple and multiple regression.

f) Test the hypothesis that ED has no independent effect on hourly wage, ie
ED coefficient=0 at the 5% significance level.

4. a) Using the data in the file estimate the linear relationship between
hourly wage, AGE, MARR, FE, and MARR*FE (see Ass2 for instructions on
creating MARR*FE).

b) Explain your results (interpret the coefficients).

c) Test the individual significance of the constant and each of the
coefficients.

5. a) Using the data in the file estimate the relationship between hourly
wage, AGE, MARR, FE, MARR*FE, and SOUTH.

b) Explain your results (interpret the coefficients).

c) Test the individual significance of the constant and each of the
coefficients.

d) Discuss the effect of the inclusion of SOUTH on the other coefficients.
Did the significance of any coefficients change? Explain why this may or
may not happen.

Return to Econ 31