I am currently doing a project based upon attitudes towards weight. I have been able to conduct repeated measures ANOVAs for my overall dataset.
However, I was further interested in how an additional factor might impact the findings (the education level the person has recieved). I thought I needed to therefore conducted a two way mixed ANOVA (since I have a within and a between factor). However, after conducting levenes test it is apparent the homogeneity is violated.
My groups (the amount of education the person has recieved) are also very different in sizes (n = 200 right down to n = 14)....so I dont think I can discount the issue with homogeneity....
I am using SPSS
What do I do next? (note - i'm not a stats person at all - speak simply please!)
It is not the original data that has the homogeneous requirement but rather the model's residuals. Check these usually with a graph. See a statistics book for details. Best, David
Have you considered using the MIXED command to estimate your model? It allows one to estimate a model that does not assume homogeneity of variance for between-subjects factors. HTH.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA