Consider this simple data file that has nine subjects (sub) in three groups (iv) with a score on the outcome or dependent variable (dv).
data list list / sub iv dv. begin data 1 1 48 2 1 49 3 1 50 4 2 17 5 2 20 6 2 23 7 3 28 8 3 30 9 3 32 end data.
Below we use the means command to find the overall mean and the means for the three groups.
means tables = dv by iv.
As we see below, the overall mean is 33, and the means for groups 1, 2 and 3 are 49, 20 and 30 respectively.
Case Processing Summary
Cases Included Excluded Total N Percent N Percent N Percent DV * IV 9 100.0% 0 .0% 9 100.0%
Report
DVIV Mean N Std. Deviation 1.00 49.0000 3 1.00000 2.00 20.0000 3 3.00000 3.00 30.0000 3 2.00000 Total 33.0000 9 12.89380
Let's run a standard ANOVA on these data using glm.
glm dv by iv.
The results of the ANOVA are shown below.
Between-Subjects Factors
N IV 1.00 3 2.00 3 3.00 3
Tests of Between-Subjects Effects
Dependent Variable: DVSource Type III Sum of Squares df Mean Square F Sig. Corrected Model 1302.000(a) 2 651.000 139.500 .000 Intercept 9801.000 1 9801.000 2100.214 .000 IV 1302.000 2 651.000 139.500 .000 Error 28.000 6 4.667
Total 11131.000 9
Corrected Total 1330.000 8
a R Squared = .979 (Adjusted R Squared = .972)
Now, let's take this information we have found and relate it to the results that we get when we run a similar analysis using dummy coding. Let's make a data file called dummy2 that has dummy variables called iv1 (1 if iv=1), iv2 (1 if iv=2) and iv3 (1 if iv=3). Note that iv3 is not really necessary, but it could be useful for further exploring the meaning of dummy variables. We will then use the regression command to predict dv from iv1 and iv2.
compute iv1 = 0. if iv = 1 iv1 = 1. compute iv2 = 0. if iv = 2 iv2 = 1. compute iv3 = 0. if iv = 3 iv3 = 1. execute. regression /dependent = dv /method = enter iv1 iv2.
The output is shown below.
Variables Entered/Removed(b) Model Variables Entered Variables Removed Method 1 IV2, IV1(a) . Enter a All requested variables entered. b Dependent Variable: DV
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .989(a) .979 .972 2.16025 a Predictors: (Constant), IV2, IV1
ANOVA(b) Model Sum of Squares df Mean Square F Sig. 1 Regression 1302.000 2 651.000 139.500 .000(a) Residual 28.000 6 4.667 Total 1330.000 8 a Predictors: (Constant), IV2, IV1 b Dependent Variable: DV
Coefficients(a) Unstandardized Coefficients Standardized Coefficients t Sig. Model B Std. Error Beta
1 (Constant) 30.000 1.247 24.054 .000 IV1 19.000 1.764 .737 10.772 .000 IV2 -10.000 1.764 -.388 -5.669 .001 a Dependent Variable: DV
First, note that from the ANOVA using the glm command that the F-value was 139.5 and for the regression using the regression command the F-value (for the model) is also 139.5. This illustrates that the overall test of the model using regression is really the same as doing an ANOVA.
After the ANOVA table, there is a table entitled Coefficients. What is the interpretation of the values listed there, the 30, 19 and -10? Notice how we have iv1 and iv2 that refer to group 1 and group 2, but we did not include any dummy variable referring to group 3. Group 3 is often called the omitted group or reference group. Recall that the means of the 3 groups were 49, 20 and 30 respectively. The intercept term is the mean of the dependent variable, which we called dv, for the omitted group, and indeed the parameter estimate (in the column B) from the output is the mean of group 3, 30. The parameter estimate for iv1 is the mean of the dependent variable, dv, for group 1 minus the mean of the dependent variable for group 3, 49 - 30 = 19, and indeed that is the parameter estimate for iv1. Likewise, the parameter estimate for iv2 is the mean of the dependent variable for group 2 minus the mean of the dependent variable for group 3, 20 - 30 = -10, the parameter estimate for iv2.
So, in summary:
Intercept mean of group 3 (mean of omitted group) iv1 mean of group 1 - group 3 (omitted group) iv2 mean of group 2 - group 3 (omitted group)
Try running this example, but use iv2 and iv3 using regression (making group 1 the omitted group) and see what happens.
Finally, consider how the parameter estimates can be used in the regression model to obtain the means for the groups (the predicted values).
The regression model is
Ypredicted = 30 + iv1*19 + iv2*-10For group 1: Ypredicted = 30 + 1 * 19 + 0 * -10 = 49 For group 2: Ypredicted = 30 + 0 * 19 + 1 * -10 = 20 For group 3: Ypredicted = 30 + 0 * 19 + 0 * -10 = 30
As you see, the regression formula predicts that each group will have the mean value of its group.
You can also perform the same analysis using glm. The print = parameter subcommand tells SPSS to print the regression coefficients.
glm dv with iv1 iv2 /print = parameter.
Tests of Between-Subjects Effects
Dependent Variable: DVSource Type III Sum of Squares df Mean Square F Sig. Corrected Model 1302.000(a) 2 651.000 139.500 .000 Intercept 2700.000 1 2700.000 578.571 .000 IV1 541.500 1 541.500 116.036 .000 IV2 150.000 1 150.000 32.143 .001 Error 28.000 6 4.667 Total 11131.000 9 Corrected Total 1330.000 8 a R Squared = .979 (Adjusted R Squared = .972)
Parameter Estimates
Dependent Variable: DVB Std. Error t Sig. 95% Confidence Interval Parameter
Lower Bound Upper Bound Intercept 30.000 1.247 24.054 .000 26.948 33.052 IV1 19.000 1.764 10.772 .000 14.684 23.316 IV2 -10.000 1.764 -5.669 .001 -14.316 -5.684