Multiple Regression Flashcards
Multiple Regression
Used to predict a continuous outcome from several categorical or continuous predictors.
Multiple Regression Statistic
R-squared is the key statistic (estimation of the amount of variability that can be accounted for by the predictor/s.
Hierarchical Regression
Experimenter decides the order of predictor entry, based on theory. Good for assessing unique influence of unknown predictors on the outcome, however requires a very good knowledge of the research area.
Forced Entry Regression
All predictors are entered simultaneously. Theory driven, however the results depend on the variables entered.
Stepwise Regression
Predictors are mathematically selected by SPSS to go into the model. Should only be used for exploratory research because it is influenced by the correlations between the variables.
Semi-partial Correlation
Measures the relationship between two variables while controlling for the effect of a third variable on only one of the original two variables.
Beta Values in Regression
Beta (b) reflects the change in the outcome associated with a unit change in the predictor. Standardised Beta (B) reflects the change in the outcome associated with a SD change in the predictor.
Assessing Model Fit of Multiple Regression
Standardized residuals - indicate outliers (95% should fall between -2 and +2).
Cook’s Distance - indicates the influence of a single case on the data (value over 1 is problematic)
Assumptions of Multiple Regression
Normality, Homogeneity of variance, linearity, independence, multicollinearity, homoscedacity, independent errors.