Tutorial Chapter 5: Multiple Regression Analysis
1) Consider the following model:
Suppose you leave out the years of experience variable. What kinds of problems or biases
would you expect? Explain verbally.
(X2 is the coefficient of education and X3 is the years of experience)
In this situation, when one variable from the model is left out, the omitted variable bias is
expected. The nature of the bias here is connected with the correlation between X2 and
X3 (X2 is the coefficient of education and X3 is the years of experience). Therefore, if the
"years of experience" variable is left out of the model, the education coefficient is biased.
2) Suppose you estimate the consumption function:
Yi = α1 + α2Xi + µ1i
And the savings function:
Zi = β1 + β2Xi + µ2i
Where Y = consumption, Z = savings, X = income and X = Y + Z, that is, income is equal to
consumption plus savings.
Based on the above statements, what is the relationship, if any, between α2 and β2? Show
your calculations.
X = (α1 + α2Xi + µ1i) + (β1 + β2Xi + µ2i)
X = α1 + β1 + α2Xi + β2Xi + µ1i + µ2i
3) Table 5.1 gives quarterly data on these variables:
Y = quantity of roses sold, dozens
X2 = average wholesale price of roses, $/dozen
X3 = average wholesale price of carnations, $/dozen
X4 = average weekly family disposable income, $/week
X5 = the trend variable taking values of 1,2 and so on, for the period 1971 Q3 to 1975 Q2 in
the Detroit metropolitan area.
You are asked to consider the following demand functions:
Yt = α1 + α2X2t + α3X3t + α4X4t + α5X5t + µt
lnYt = β1 + β2 lnX2t + β3 lnX3t + β4 lnX4t + β5 lnX5t + µt
a) Estimate the parameters of the linear model and interpret the results.
Dependent Variable: Y
Method: Least Squares
Date: 01/07/21 Time: 22:23
Sample: 1971Q3 1975Q2
Included observations: 16
Variable Coefficient Std. Error t-Statistic Prob.
X2 -2227.704 920.4657 -2.420193 0.0340
X3 1251.141 1157.021 1.081347 0.3027
X4 6.282986 30.62166 0.205181 0.8412
X5 -197.3999 101.5612 -1.943655 0.0780
C 10816.04 5988.348 1.806181 0.0983
R-squared 0.834699 Mean dependent var 7645.000
Adjusted R-squared 0.774590 S.D. dependent var 2042.814
S.E. of regression 969.8744 Akaike info criterion 16.84252
Sum squared resid 10347220 Schwarz criterion 17.08395
Log likelihood -129.7401 Hannan-Quinn criter. 16.85488
F-statistic 13.88635 Durbin-Watson stat 2.333986
Prob(F-statistic) 0.000281
b) Estimate the parameters of the log-linear model and interpret the results.
Dependent Variable: LNY
Method: Least Squares
Date: 01/07/21 Time: 22:29
Sample: 1971Q3 1975Q2
Included observations: 16
Variable Coefficient Std. Error t-Statistic Prob.
LNX2 -1.273555 0.526649 -2.418224 0.0341
LNX3 0.937305 0.659191 1.421902 0.1828
LNX4 1.712976 1.200843 1.426478 0.1815
LNX5 -0.181597 0.127893 -1.419907 0.1833
C 0.626824 6.148262 0.101951 0.9206
R-squared 0.777953 Mean dependent var 8.902209
Adjusted R-squared 0.697208 S.D. dependent var 0.306877
S.E. of regression 0.168864 Akaike info criterion -0.469145
Sum squared resid 0.313664 Schwarz criterion -0.227711
Log likelihood 8.753157 Hannan-Quinn criter. -0.456781
F-statistic 9.634745 Durbin-Watson stat 1.782659
Prob(F-statistic) 0.001343
