Finite population correction

Discuss statistics related things

by dsrtrosy » Wed Aug 07, 2019 7:12 pm

I am no statistician, but my advisor wants me to take some data that lends itself to yes/no answers and run a logistic regression. I worked with a statistician to set up the research design, but she never suggested that the small sample size would be a problem considering the extremely small size of the population. However, having run the regression in SAS, it clearly was a problem.

A colleague suggested using a finite population correction factor. I'm not sure what that is (mathematically) or how to use it, but it sounds like EXACTLY what I need. Is this something I can do in jamovi? Is anyone using something like that?

Any help or other friendly suggestions greatly appreciated!
dsrtrosy
 
Posts: 2
Joined: Wed Aug 07, 2019 7:06 pm

by jonathon » Thu Aug 08, 2019 6:28 am

hi,

sorry, i'm pretty sure we don't have that. the best way forward may be to dabble in R. you could use the Rj editor for that (then you can do it inside jamovi).

sorry i can't be more help

jonathon
User avatar
jonathon
 
Posts: 899
Joined: Fri Jan 27, 2017 10:04 am

by dsrtrosy » Thu Aug 08, 2019 4:32 pm

Thanks, you actually were helpful. I have only barely dabbled in R! But that might work.
dsrtrosy
 
Posts: 2
Joined: Wed Aug 07, 2019 7:06 pm


Return to Statistics