Multiple Regression
- Created by: Chloe
- Created on: 30-04-15 15:24
View mindmap
- Multiple Regression
- The prediction of the DV from a set of predictors, there are two regression lines for each predictor
- A is the intercept on the Y axis when X=0, but there are different B values for each predictor (the relationship of each predictor to the outcome variable is different)
- The Multiple Coefficient of Determination: the best measure of how well the regression equation fits the data
- R^2 - The square of the multiple correlation R.
- The value of R^2 is adjusted (reduced to compensate for the fact that it tends to increase with the numbers of predictors & to compensate for the low reliability of small sample sizes
- R^2 - The square of the multiple correlation R.
- Significance of prediction: he overall MR analysis tests the null hypothesis that none of the predictors has any effect on the outcome variable
- Comparison of MLR methods:
- Enter: all variables are entered into the equation, irrespective of their degree of overlap or predictive value
- Stepwise: variables are entered in the equation in the order of their correlation with the criterion - the IV with the highest R value first and so on - for the second and subsequent IV''S the correlation is controlled for those already entered (it is the partial correlation that is important)
- Hierarchical: MLR is used to test theory rather than produce regression equations - (1) first variables entered are demographic or background variables (age, gender etc) that the investigator wishes to control for (2) then variables are entered in turn, each controlling for all those entered in earlier steps
- The prediction of the DV from a set of predictors, there are two regression lines for each predictor
Comments
No comments have yet been made