Scribd
Upload
9 views3 pages

Problem Set - Chapter 3

The document outlines a series of exercises focused on regression analysis using various Excel data sets, including EXECUTIVE.XLS and others. Each exercise involves running regressions, interpreting results, calculating confidence intervals, and testing hypotheses related to the relationships between different variables. The exercises aim to enhance understanding of regression techniques, residual analysis, and the implications of statistical findings in the context of executive compensation and market capitalization.

Uploaded by

Tuan Anh Tran
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
9 views3 pages

Problem Set - Chapter 3

The document outlines a series of exercises focused on regression analysis using various Excel data sets, including EXECUTIVE.XLS and others. Each exercise involves running regressions, interpreting results, calculating confidence intervals, and testing hypotheses related to the relationships between different variables. The exercises aim to enhance understanding of regression techniques, residual analysis, and the implications of statistical findings in the context of executive compensation and market capitalization.

Uploaded by

Tuan Anh Tran
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Chapter 3: An introduction to simple regression

Exercise 1
The Excel data set EXECUTIVE.XLS contains data on Y = executive
compensation, X = profits, W = percentage change in sales and Z = percentage
change in debt.
(a) Run a regression of Y on X and interpret the results.
(b) Run a regression of Y on W and one of Y on Z and interpret the results.
(c) Create a new variable, V, by dividing X by 100. What are the units in terms of
which V is measured?
(d) Run a regression of Y on V. Compare your results to those for (a). How do you
interpret your coefficient estimate of ? How does α^ differ between a) and d)?
(e) Experiment with scaling dependent and explanatory variables (i.e. by dividing
them by a constant) and see what effect this has on your coefficient estimates.
Exercise 2
Using the data in EXECUTIVE.XLS (see Exercise 1), run a regression of Y on X
and graphically and numerically compare actual and fitted values. In Excel this can
be done by running the regression with the box clicked on labeled “Line Fit Plot”
in the Regression menu (i.e. look at the columns labeled “Residual Output” and the
accompanying display chart).
Exercise 3
(a) Using the data in EXECUTIVE.XLS (see Exercise 1) run a regression of Yon
X and graphically and numerically compare residuals. In Excel this can be done
with the boxes labeled “Residuals” and “Residual Plots” in the regression menu
clicked on. How would you interpret the residuals? Are there any outliers?
(b) Repeat question (a) for the other variables, W and Z in this data set.
Exercise 4
(a) Using the data in EXECUTIVE.XLS (see Exercise 1) run a regression of Y on
X using Excel. What is the R2?
(b) Calculate the correlation between Y and X.
(c) Discuss the relationship between your answers in (a) and (b).
(d) Redo (a) for various regressions involving the variables W, X, Y and Z in
the data set. Comment on the fit of each of these regressions.
Exercise 5
The data sets used to calculate Figures 5.1, 5.2, 5.3 and 5.4 are in FIG51.XLS,
FIG52.XLS, FIG53.XLS and FIG54.XLS.
(a) Calculate the OLS estimates and for these four data sets. How close are they to
0 and 1 (the values we used to artificially simulate the data)?
(b) Calculate confidence intervals for a for the four data sets. Examine how the
width of the confidence interval relates to N and the variability of the errors.
(c) Calculate 99% and 90% confidence intervals for the data sets. How do these
differ from the 95% confidence intervals in (b)?
Exercise 6
The file EQUITY.XLS contains data on Y = market capitalization and X = debt
(both measured in millions of dollars) for N = 309 companies in the USA.
(a) Run a regression of Y on X and obtain 95% confidence intervals for  and 
(b) Write a sentence explaining verbally what the 95% confidence interval for 
means in terms of the possible range of values that the effect of the explanatory
variable on the dependent variable may take.
(c) The file EQUITY.XLS also contains data on other potential explanatory
variables: sales, income and assets (all measured in millions of dollars). Repeat
part (a) and (b) using these explanatory variables one at a time.
Exercise 7
Running a regression using data set EXECUTIVE.XLS and test the hypothesis that
 = 0.
Exercise 8
The Excel data set EXECUTIVE.XLS contains data on Y = executive
compensation, X = profits, W = change in sales and Z = change in debt.
(a) Run a regression of Y on W and interpret your results. Can you reject the
hypothesis that changes in sales has an effect on executive compensation?
(b) Run a regression of Y on Z and interpret your results. Can you reject the
hypothesis that changes in debt have an effect on executive compensation?
Exercise 9
Use data sets FIG51.XLS, FIG52.XLS, FIG53.XLS and FIG54.XLS.
(a) Test whether  = 0 using the confidence interval approach for each of the four
data sets
(b) Test whether  = 0 using the P-value approach and the four data sets. Use the
5% level of significance.
(c) Redo (a) and (b) for .
(d) Redo parts (a), (b) and (c) using the 1% level of significance.
(e) Are your results sensible in light of the discussion in this chapter of the factors
affecting the accuracy of OLS estimates?
Exercise 10
The file ADVERT.XLS contains data on the sales and advertising expenditures of
84 companies. Set up and run a regression using this data and discuss your results
verbally as you would in a report. Include a discussion of the marginal effect of
advertising on sales and a discussion of whether this marginal effect is statistically
significant.
Exercise 11
Use data sets FIG51.XLS, FIG52.XLS, FIG53.XLS and FIG54.XLS.
Test whether R2 = 0 for each of the four data sets. Compare your results with those
of Exercise 9.

You might also like