search

cambiotraining / corestats

These materials support the Core statistics sessions, run through the Bioinformatics Training Facility at the University of Cambridge.
https://cambiotraining.github.io/corestats/
MIT License
3 stars 6 forks source link

section 11.8 KW lobster revisited #14

Open mvanrongen opened 1 year ago

mvanrongen commented 1 year ago

section 11.8 KW lobster revisited Also can I suggest swapping the Lobsters(revisited) example out for something else, given that we concluded that the appropriate thing to do is use the ANOVA not the KW?

mvanrongen commented 1 year ago

an error in the course materials in section 14.8.5. The interpretation is mixing up Tulips and Auxin.

mvanrongen commented 1 year ago

Just going through the Clover & Yarrow example, which ends with the model: log(yield) ~ yarrow * farm

But the interaction is not significant. So it should be removed. And it looks like in the model: log(yield) ~ yarrow + farm the Farm term wouldn't be significant either

So perhaps the example should end with the model: log(yield) ~ yarrow

I think it may confuse the students to leave them with a model that includes terms that shouldn't be there - it's clear that Farm has no effect in this case