Hello - last minute question!
Regarding assumption 5 and homoscedasticity, I think I understand that the rule of thumb is that the assumption is violated when the highest variance is 10 times bigger than the lowest variance, ie. with a ratio of 10:1. However, on the "Assumptions in Regression" handout, the Heteroscedasticity Example does not have a variance 10 times bigger than the lowest varience, but there is heteroscedasticity and assumption 5 is violated. Could you clarify? Thank you!
Subscribe to:
Post Comments (Atom)
4 comments:
Hello!
We think that you misunderstood something on the handout. In order to check the ratio of variance, a specific calculation needs to be done. You cannot guess this ratio from looking at the graph. The range of the results that can be seen on the graph does not equal variance.
Rule of thumb still applies.
~(andy)
Ok - Could you provide an example of how to check/ calculate the ratio of variance and review how to determine the rule of thumb? Thanks!
Before you are make a plot with residuals, you have to create a new variable for residuals on spss, right? Look at the residuals on the new column. Find out the largest residual and divide it by the smallest residual, then, you will get the ratio of variance!
Post a Comment