Complete Smart Alex's Task #1 attached to perform a multiple regression analysis using the Supermodel.sav dataset attached.
- State underlying assumptions
- Determine whether assumptions have been met
- Propose alternatives if assumptions are not met
- State null and alternative hypotheses
- Analyze data using IBM SPSS Statistics
- Interpret and report the results with IBM SPSS Statistics, including effect size
- Describe sample size
Construct a theoretical framework to support the research proposal
Multiple Regression Analysis Smart Alex Supermodel data set
State underlying assumptions
1. Variable have a normal distribution
2. There is a linear relationship between the independent and dependent variables
3. Variance across independent variables is similar (HOMOSCEDASTICITY)
4. There are no serious outliers that will "pull" the model and distort the relationship
Determine whether assumptions have been met
The skewness and kutosis for the independent variables are within 1 and -1 so they look normal enough. The normal P-P plot, however shows that the data doesn't "hug" the line very well indicating that normality assumption is stretched a bit. Regression is not too sensitive to lack of normality so we are probably ok. The scatterplot of predicted values versus residual values (ZPRED vs.ZRESID) gets wider as you more to the right indicating that the variance is not similar across the variables. Finally, there are several fairly extreme outliers over three standard deviations. Having one or two data points near three standard deviations out of a sample of 231 might be ...
Your tutorial is 642 words plus four references and gives you an outline that responds to each point and an SPSS output file showing the results.