Regression Analyses

• Normality of data is not required.
• Used when correlation is significant (Baron and Kenny method)
• Method
  • Click on Analyse
  • Regression
  • Linear
    • Method: Enter/Stepwise
      • Enter-All variables considered; for testing existing well-established theories.
      • Step-wise-only significant variables used; for developing/ exploring new theories.
  • Shift IV and DV
  • Ok

• Output
  • Check F-value and its significance. If significant then look at R² score.
    • F-value in the ANOVA table indicates the fit of the scores for linear regression.
      • If F value’s significance is less than 0.05, then the data fits the model.
• For the research (sample)-R²
• For population-adjusted R²
  • The R² indicates how much variation in DV can be explained by IV. In this case, .414 means 41% of the change in self-esteem can be predicted by self-confidence, i.e., the contribution of IV to DV
• Check Beta value
  • The unstandardized coefficient is used
• Prediction can be done with unstandardized coefficient
• Significance level should be less than 0.05
• Beta values help to predict exact change in unit.
  • B (Constant+IV)