The McGraw-Hill Series Economics essentials of economics brue, McConnell, and Flynn Essentials of Economics

Summary and 
Conclusions
304
Part One
Single-Equation Regression Models
1. Dummy variables, taking values of 1 and zero (or their linear transforms), are a means
of introducing qualitative regressors in regression models.
2. Dummy variables are a data-classifying device in that they divide a sample into various
subgroups based on qualities or attributes (gender, marital status, race, religion, etc.)
and 
implicitly 
allow one to run individual regressions for each subgroup. If there are
differences in the response of the regressand to the variation in the qualitative variables
in the various subgroups, they will be reflected in the differences in the intercepts or
slope coefficients, or both, of the various subgroup regressions.
3. Although a versatile tool, the dummy variable technique needs to be handled carefully.
First,
if the regression contains a constant term, the number of dummy variables must be
one less than the number of classifications of each qualitative variable. 
Second,
the
coefficient attached to the dummy variables must 
always
be interpreted in relation to
the base, or reference, group—that is, the group that receives the value of zero. The base
chosen will depend on the purpose of research at hand. 
Finally,
if a model has several
qualitative variables with several classes, introduction of dummy variables can consume
a large number of degrees of freedom. Therefore, one should always weigh the number
of dummy variables to be introduced against the total number of observations available
for analysis.
Dependent Variable: LOG(WI)
Method: Least Squares
Sample: 1 261
Included observations: 261
Coefficient
Std. Error
t
-Statistic
Prob.
C
3.836483
0.106785
35.92725
0.0000
AGE
0.025990
0.003170
8.197991
0.0000
D
SEX
-0.868617
0.106429
-8.161508
0.0000
D
SEX
*
DE
2
0.200823
0.259511
0.773851
0.4397
D
SEX
*
DE
3
0.716722
0.245021
2.925140
0.0038
D
SEX
*
DE
4
0.752652
0.265975
2.829789
0.0050
DPT
0.627272
0.078869
7.953332
0.0000
R-squared
0.514449
Mean dependent var.
4.793390
Adjusted R-squared
0.502979
S.D. dependent var.
0.834277
S.E. of regression
0.588163
Akaike info criterion
1.802828
Sum squared resid.
87.86766
Schwarz criterion
1.898429
Log likelihood
-228.2691
Hannan-Quinn criter.
1.841257
F
-statistic
44.85284
Durbin-Watson stat.
1.873421
Prob (
F
-statistic)
0.000000 
Interestingly, if we drop the education dummies but retain the interaction dummies, we
obtain the following results:
guj75772_ch09.qxd 12/08/2008 04:19 PM Page 304

---

Chapter 9
Dummy Variable Regression Models