16 - Regression And Prediction

Question 1

Regression

Answer 1

Allows the researcher to make predictions of the likely values of the dependent variable (Y) from know values of the independent variable (X)

F ratio should be significant.

Question 2

Regression assumptions

Answer 2

• minimum requirement: at least 15*(nb IV)
• identify and remove outliers
• normality of residuals
• homoscedasticity
• linear dispersion of points
• avoid multicollinearity (multiple regression)

Question 3

Line of best fit

Answer 3

Regression equation will minimise the sum of square of the vertical distances between the actual data point and the line and therefore make error as small as possible.

Y= a +bX + error

Y= criterion
X= predictor

Question 4

Standard error of the estimate (regression)

Answer 4

SE= SD*sqrd(1-r^2)

Question 5

r^2 and R^2

Answer 5

Simple

Multiple regression

Question 6

VIF

Answer 6

The variance inflation factor (VIF) measures the impact of collinearity among the IVs in a multiple regression model on the estimation. It expresses the degree to which collinearity among the predictors dégradés the precision of an estimate. Concern if VIF>10

Question 7

Hierarchical multiple regression

Answer 7

The researcher determines the order of entry of the IVs into the equation.

Question 8

Stepwise multiple regression

Answer 8

• forward entry: one at a time

- backward entry: all at the same time