Does a Weighted Paired T-test exist?
I am comparing the percent of a population that are juvenile between 'Autumn' and 'Spring' at many sites. The Sample size (n) is variable among the sites and between the seasons. Sample data:
Site___Fall %__n_____Spring %__n
A_____68%___50_____60%_____300
B_____81%___200____68%_____30
C_____77%___1000___72%_____25
D...
If samples were equal I would use a paired T-test but since they are not, I would prefer to use weighted means. Sample size would be used as weight because I have greater confidence in them. I need to test non-independent weighted means but I am not aware of a 'Weighted Paired T-test.' What statistical test do you recommend?
Thank you
p.s. people have suggested using linear regression for this, but I don't know how or if it's even right.
I am comparing the percent of a population that are juvenile between 'Autumn' and 'Spring' at many sites. The Sample size (n) is variable among the sites and between the seasons. Sample data:
Site___Fall %__n_____Spring %__n
A_____68%___50_____60%_____300
B_____81%___200____68%_____30
C_____77%___1000___72%_____25
D...
If samples were equal I would use a paired T-test but since they are not, I would prefer to use weighted means. Sample size would be used as weight because I have greater confidence in them. I need to test non-independent weighted means but I am not aware of a 'Weighted Paired T-test.' What statistical test do you recommend?
Thank you
p.s. people have suggested using linear regression for this, but I don't know how or if it's even right.
Last edited: