Multiple Regression

Multiple Regression

Measures the rela between the dependent variable (Y) and more than one independent variable

Regression analysis assumes that a variable’s variance is composed of 3 components:

Common

Specific

Error

The regression line represents ______

the best summary of the linear relationship between the variables

Multiple regression equation:

Y’ = a + b1×1 + b2×2 + ….

b1 = slope of first IV

x1 = value of second IV

Standardized Beta

Compares the strength of the IVs on the same standardized scale

Independent effect

Multicollinearity

Multicollinearity: Two or more predictor variables in a multiple regression model are highly correlated

Suppression

Regression model improves by adding a variable uncorrelated with the dependent variable, but related to the independent variable

Limitations to multiple regression: Researcher Bias

Researcher may enter variables in a way that supports their theories

Limitations to multiple regression: Overfitting

More independent variables in a regression results in additional variance that can be explained

Adjusted R squared

Regressions with different numbers of IVs can be compared

Theory

Dictates which variables should be included or excluded from the analysis