beta p-value SE

Beta value

Beta值（β） effect size 线性回归的参数https://www.mv.helsinki.fi/home/mjxpirin/GWAS_course/material/GWAS2.html

Hello Mohsin,

 

Cohen's f-squared would reflect the explanatory power of the overall regression model: R-squared (the explained variance) divided by (1 - R-squared) (the unexplained variance). In the case of a single predictor model, beta (the standardized regression coefficient) = r(x,y) = R, so beta would be related to f-squared as:

 

f-squared = beta-squared / (1 - beta-squared),

 

beta-squared = f-squared / (1 + f-squared), and

 

beta = square root of [f-squared / (1 + f-squared)].

 

R-squared, f-squared, and beta can and have been used as effect size indicators. A common question is, are they (sufficiently) different from zero to be considered noteworthy?

 

Good luck with your work!

 

I am assuming you are speaking of teh coefficients in linear regression. When your response variable is metric and can readily be interpreted in terms of impact, the beta coefficients are effects sizes by themselve

 

http://www.iikx.com/news/statistics/1827.html

 

 

 

 

 

SE: standard error

 

 

SE = sqrt[SD1/n1 + SD2/n2]

P-value ≈ [Effect size/SE]

SE depends on N; the bigger the N the smaller the SE

P-value depends on SE and effect size; the bigger the SE the higher the P-value

Thus, P-value depends on N; the bigger the N the lower the P-value

Therefore, we can say in a GWAS that a lower P-value indicates a smaller SE or a higher effect size

 

 

https://www.mv.helsinki.fi/home/mjxpirin/GWAS_course/material/GWAS2.html