My ANOVA has total N of 457 and 7 groups within this. Equality of variances not met so I was going to use games howell. I've heard with large groups equality of variances is important, but normality isn't as important so a parametric test like the ANOVA should still be okay. Is this correct?
It seems like you're dealing with a situation where the assumption of homogeneity of variances (homoscedasticity) is not met for your ANOVA, and you are considering using the Games-Howell post hoc test due to this violation. Let's address your points:
To summarize, if you have violated the assumption of homogeneity of variances, using Games-Howell as a post hoc test is a reasonable approach. As long as your sample sizes are reasonably large, the violation of normality might not be a critical concern for the ANOVA results. Just be aware that ANOVA is more robust to normality violations when sample sizes are larger. Always consider the context of your data, the research question, and the potential impact of assumptions on your conclusions. If in doubt, it's a good idea to consult with a statistician or someone experienced in data analysis for your specific field.
Yes I'm using Games-Howell but normality is also not met. But I've read that if you have a large sample size this can be overlooked instead of using a non-parametric test. Is this correct?
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping
- Maps