Ceiling effects in regression

#1
Hi all,

I have quite a marked ceiling effect in my regression - what will this do in terms of assumptions? is it just normality or also homeodascity that will be violated?

How is best to proceed given this? transformation does not reduce it :-\
 

TheEcologist

Global Moderator
#2
Hi all,

I have quite a marked ceiling effect in my regression - what will this do in terms of assumptions? is it just normality or also homeodascity that will be violated?

How is best to proceed given this? transformation does not reduce it :-\
Post a scatter plot of your data, we can better help then.

If you have a strong asymptotic relationship in your data, why don’t you try a logistic curve? Or maybe a Michealis-Menten curve?

Anyhow a scatter plot should give us a beter idea of the scatter.