Multiple Regression

1
Multiple Regression

• What is our research question?
• How does IV1, IV2, IV3 influence the DV
• What unique contribution to variation in the DV
  does IV1 make when holding IV2 IV3 constant?
• What unique contribution to the DV does IV2 make
  when holding IV1 IV3 constant?

2

• Example
• TVhours regressed onto Educ, Age, Income
• Hr1 As education increases TV viewing
  decreases.
• Hr2 As income increases TV viewing
  decreases.
• Hr3 As age increases TV viewing increases.

3
Unique Shared Variation
4
Output

• Important Statistics
• R-squared The amount of variation
  explained by all of the IVs taken
  together.
• AVOVA Box
• F Test Tells us if the regression model is a
  good fit for this data.
• T-Test Significance test for each Beta.

5

• How to read B Beta
• B is the unstandardized beta.
• Tells us the amount of change in the DV for a one
  unite change an IV, holding all other IVs
  constant.
• Beta is the standardized beta.
• Tells us about the strength of an IV on the DV
  holding all other IVs constant.
• The largest Beta has the strongest impact on the
  IV. Beta can range from -1 to 1 with 0
  indicating no relationship.

6
SPSS Output

• See overhead