Regression
(I have provided additional information about regression for those who are interested. This is not required material for EPSY 5601)
SPSS Printout
Variables Entered/Removed
|
Model |
Variables Entered |
Variables Removed |
Method |
|
1 |
Educational level (years) |
. |
Stepwise (Criteria: Probability-of-F-to-enter <= .050, Probability-of-F-to-remove >= .100). |
|
2 |
Gender |
. |
Stepwise (Criteria: Probability-of-F-to-enter <= .050, Probability-of-F-to-remove >= .100). |
|
3 |
Previous experience (months) |
. |
Stepwise (Criteria: Probability-of-F-to-enter <= .050, Probability-of-F-to-remove >= .100). |
a Dependent Variable: Beginning salary
Model Summary
|
R |
R Square |
Adjusted R Square |
Std. Error of the Estimate |
Change Statistics |
|||||
|
Model |
R Square Change |
F Change |
df1 |
df2 |
Sig. F Change |
||||
|
1 |
.633 |
.401 |
.400 |
$6,098.26 |
.401 |
315.897 |
1 |
472 |
.000 |
|
2 |
.680 |
.462 |
.460 |
$5,784.26 |
.061 |
53.637 |
1 |
471 |
.000 |
|
3 |
.696 |
.484 |
.481 |
$5,671.16 |
.022 |
19.974 |
1 |
470 |
.000 |
a Predictors: (Constant), Educational level (years)
b Predictors: (Constant), Educational level (years), Gender
c Predictors: (Constant), Educational level (years), Gender, Previous experience (months)
ANOVA
|
Model |
Sum of Squares |
df |
Mean Square |
F |
Sig. |
|
|
1 |
Regression |
11747808912.317 |
1 |
11747808912.317 |
315.897 |
.000 |
|
Residual |
17553096053.137 |
472 |
37188762.824 |
|||
|
Total |
29300904965.454 |
473 |
||||
|
2 |
Regression |
13542369102.880 |
2 |
6771184551.440 |
202.381 |
.000 |
|
Residual |
15758535862.574 |
471 |
33457613.296 |
|||
|
Total |
29300904965.454 |
473 |
||||
|
3 |
Regression |
14184764846.649 |
3 |
4728254948.883 |
147.014 |
.000 |
|
Residual |
15116140118.804 |
470 |
32162000.253 |
|||
|
Total |
29300904965.454 |
473 |
a Predictors: (Constant), Educational level (years)
b Predictors: (Constant), Educational level (years), Gender
c Predictors: (Constant), Educational level (years), Gender, Previous experience (months)
d Dependent Variable: Beginning salary
Coefficients
|
Unstandardized Coefficients |
Standardized Coefficients |
t |
Sig. |
|||
|
Model |
B |
Std. Error |
Beta |
|||
|
1 |
(Constant) |
-6290.967 |
1340.920 |
-4.692 |
.000 |
|
|
Educational level (years) |
1727.528 |
97.197 |
.633 |
17.773 |
.000 |
|
|
2 |
(Constant) |
-5096.451 |
1282.290 |
-3.974 |
.000 |
|
|
Educational level (years) |
1470.321 |
98.655 |
.539 |
14.904 |
.000 |
|
|
Gender |
4180.769 |
570.853 |
.265 |
7.324 |
.000 |
|
|
3 |
(Constant) |
-7938.049 |
1408.851 |
-5.634 |
.000 |
|
|
Educational level (years) |
1625.292 |
102.753 |
.596 |
15.817 |
.000 |
|
|
Gender |
3446.504 |
583.307 |
.218 |
5.909 |
.000 |
|
|
Previous experience (months) |
12.001 |
2.685 |
.159 |
4.469 |
.000 |
a Dependent Variable: Beginning salary
Excluded Variables
|
Beta In |
t |
Sig. |
Partial Correlation |
Collinearity Statistics |
||
|
Model |
Tolerance |
|||||
|
1 |
Gender |
.265 |
7.324 |
.000 |
.320 |
.873 |
|
Previous experience (months) |
.219 |
6.174 |
.000 |
.274 |
.936 |
|
|
2 |
Previous experience (months) |
.159 |
4.469 |
.000 |
.202 |
.862 |
a Predictors in the Model: (Constant), Educational level (years)
b Predictors in the Model: (Constant), Educational level (years), Gender
c Dependent Variable: Beginning salary
Interpretation of Printout:
Table 1
Summary of Stepwise Regression Analysis for Variables
Predicting Salary (N = 473)
————————————————————-
Variable B SE B b
————————————————————–
Step 1
Education Level 1727.53 97.20 .63***
Step 2
Education Level 1470.32 98.66 .54***
Gender 4180.77 570.85 .27***
Step 3
Education Level 1625.29 102.75 .60***
Gender 3446.50 583.31 .22***
Experience 12.00 2.69 .16***
————————————————————-
Note. R2 = .40 for Step 1; DR2 = .06 for Step 2;
DR2 = .02 for Step 3 (ps < .001).
***p < .001.
Forty-nine percent of the variation in beginning salary can be predicted from an employee’s educational level, gender, and previous experience.
Education level (b = .60) is the best predictor of beginning salary. It accounts for 40% (R2)of the variation in salary from one individual to another.
Education level, gender, and previous experience are statistically significant predictors.
The unstandardized regression equation is: y = -7938.05 + 1625.29 (years of education) + 3446.50 (gender) + 12.00 (months of previous experience)
The statistical significance of the model is F (3, 470) = 147.01, p < .001.
The best estimate of a beginning salary for a male (1) with 15 years of education and 150 months of experience would be -7938.05 + 1625.29 (15) + 3446.50 (1) + 12.00 (150) = $21,687.80