I want to run a one-way ANOVA on some questionnaire data in SPSS, but some of my data have unequal variance. Would it be valid to log transform the data to normalize the distribution, and then run the ANOVA? Should I just rely on a non-parametric test like Kruskall-Wallis? Thanks!
Can I log transform data to run One-Way ANOVA?
Try the log transform. The log transforms can be very useful to help in solving problems with assumptions in the statistical tests.
try transforming only the response variable, then try just the predictive variables and then both and see what happens to the models.
there is nothing wrong with using such a transform.
using the non-parametric tests can help as well, especially if their results mirror the results of the parametric test conducted.
No comments:
Post a Comment