Multiple Regression

Main aspects

  • Expresses one variable as a function of others:

    Q1 = A + B x Q2 + C x Q3 + ...

    Q1 is the dependent variable
    Q2, Q3, ... are the independent variables
    A is a constant
    B, C, ... are coefficients

  • A 'stepwise' procedure selects the best set of independent variables
  • Size and sign of coefficients are a measure of effect
  • Category-type dependent variables handled by logistic regression