step

Karabiner

TS Contributor
#2
It would be useful if you gave some more details about your study (at least topic, research questions, sample size, variables)
and what you actually mean by "publish".

With kind regards

Karabiner
 

Dason

Ambassador to the humans
#3
Also we don't live in your brain. So "(6,58,220 and 299, with Bonferoni corrections of course)" just sounds like gibberish since you provided no explanation for it.
 
#10
Federal is moot unless you are disseminating results across state lines.

Stepwise is typically frowned upon since it does not take into account the relationships between variables:

https://ftp.cs.ucla.edu/pub/stat_ser/r493.pdf
thanks a lot. but the point is that I'm not too concerned about the relationships between variables - that's not a problem. if they exist, I include them in the model with the appropriate notation (X:Y) if they matter. I'm wondering if I can do an analysis of the intermediate stages on the way from involving 6 predictors to the maximum 299 (that's 58 and 220)? (check the file)
 
Last edited: