# (solution) Assignments Applied Regression Analysis and Other Multivariable

Assignments

Applied Regression Analysis and Other Multivariable Methods, 5th ed, Kleinbaum et. al.

Incorporate any generated computer output.  Show all work.

Chapter 11:   Confounding and Interaction                                   Due ________________

Pages: 242?256           2, 3, 4

Problem 2:

Using the appropriate regression coefficient as your measure of association, determine whether confounding exists.  Explain.

Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients.  What conclusion would you draw?

Problem 3:

Using the appropriate regression coefficient as your measure of association, determine whether confounding exists.  Explain.

Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients.  What conclusion would you draw?

What does this example illustrate about using a test of the hypothesis H0: b2 = 0 in order to assess confounding.

Problem 4:

Fill in the following ANOVA table for the model .

Test if X2 should be entered into a model already containing X1.

Test if X3 should be entered into a model already containing X1 & X2.

Test if X2 & X3 should be entered into a model already containing X1.

Based upon the previous tests, what is the most appropriate regression model?

Based upon the information provided, can you assess whether X1 is a confounder of the X2?Y relationship?  Explain.

Assignments

Applied Regression Analysis and Other Multivariable Methods, 5th ed, Kleinbaum et. al.

Incorporate any generated computer output.  Show all work.

Chapter 11:   Confounding and Interaction                                   Due ________________

Pages: 242?256           2, 3, 4

Problem 2:

Using the appropriate regression coefficient as your measure of association, determine whether confounding exists.  Explain.

Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients.  What conclusion would you draw?

Problem 3:

Using the appropriate regression coefficient as your measure of association, determine whether confounding exists.  Explain.

Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients.  What conclusion would you draw?

What does this example illustrate about using a test of the hypothesis H0: b2 = 0 in order to assess confounding.

Problem 4:

Fill in the following ANOVA table for the model .

Test if X2 should be entered into a model already containing X1.

Test if X3 should be entered into a model already containing X1 & X2.

Test if X2 & X3 should be entered into a model already containing X1.

Based upon the previous tests, what is the most appropriate regression model?

Based upon the information provided, can you assess whether X1 is a confounder of the X2?Y relationship?  Explain.

Assignments

Applied Regression Analysis and Other Multivariable Methods, 5th ed, Kleinbaum et. al.

Incorporate any generated computer output.  Show all work.

Chapter 11:   Confounding and Interaction                                   Due ________________

Pages: 242?256           2, 3, 4

Problem 2:

Using the appropriate regression coefficient as your measure of association, determine whether confounding exists.  Explain.

Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients.  What conclusion would you draw?

Problem 3:

Using the appropriate regression coefficient as your measure of association, determine whether confounding exists.  Explain.

Suppose that confounding was defined to require a comparison of crude versus adjusted (partial) correlation coefficients.  What conclusion would you draw?

What does this example illustrate about using a test of the hypothesis H0: b2 = 0 in order to assess confounding.

Problem 4:

Fill in the following ANOVA table for the model .

Test if X2 should be entered into a model already containing X1.

Test if X3 should be entered into a model already containing X1 & X2.

Test if X2 & X3 should be entered into a model already containing X1.

Based upon the previous tests, what is the most appropriate regression model?

Based upon the information provided, can you assess whether X1 is a confounder of the X2?Y relationship?  Explain.

NetSales 131.634 165.1113 191.329 211.199
Gross Pro?t 28.9119 35.349 41.1114 46.23? Net Income 4.4311 5.311 6.295 6.61'1
Total Assets 49.996 111.349 13.1311 33.52?
Shareholders' Equity 21:112… 