I am analyzing a set of data of two factors, one at three and other at seven levels, to check how they influence my response variable. However, when testing the ANOVA assumpions it results it follows a normal distribution but the data is heterocedastic. I know that for heterocedastic one-way ANOVA there is the Welch test, but I did not find any alternative for the multifactor ANOVA.
Asked
Active
Viewed 26 times
0
-
You really need to tell us some more about your data (what does the response variable represent in real life?), but inn the meantime see https://stats.stackexchange.com/questions/125221/dealing-with-heteroscedasticity-in-anova, https://stats.stackexchange.com/questions/179813/what-is-an-appropriate-test-for-a-normally-distributed-heteroscedastic-multi-f, https://stats.stackexchange.com/questions/131401/how-to-get-anova-table-with-robust-standard-errors – kjetil b halvorsen May 31 '23 at 20:45