correcting raw score for age and gender

Posted on Mon, 11/29/2021 - 10:39

valentinav Joined: 06/15/2020

Forums

Hi Everyone,
I received suggestions to correct the the raw score for age and sex main effects, and for the age x sex interaction. However, I am not familiar with this procedure, does someone knows more about it?

So far, what I had done was to insert age a covariate in my analysis.

Thanks!

Replied on Fri, 02/11/2022 - 10:50

AdminNeale Joined: 03/01/2013

Can use lm() with predict

I snagged this from https://stackoverflow.com/questions/20907583/how-to-return-predicted-values-residuals-r-square-from-lm
which is ok, but to get the interaction term I am told you need to use * instead of + for the independent variables (age & sex): lm(y~age*sex)

set.seed(1) n=10


y=rnorm(n)

x1=rnorm(n)

x2=rnorm(n)
lm.ft=function(y,x1,x2)

#   return(lm(y~x1+x2)$coef)

    return(lm(y~x1+x2))

res=lm.ft(y,x1,x2) ypredicted <- predict(res) residuals <- y - ypredicted

Log in or register to post comments

Replied on Fri, 02/11/2022 - 18:54

tbates Joined: 07/31/2009

umx_residualize can help (also for twin data)

FYI, `umx_residualize()` packages this up cleanly, and works for wide data.

The help page is useful, but:

residDF = umx_residualize(mpg ~ cyl * disp, data = mtcars) plot(residDF$mpg ~mtcars$mpg)

twin example

r1 = umx_residualize(ht ~ wt, data = twinData, suffixes=1:2)

**multiple left hand side variables**
r1 = umx_residualize(var = c("disp", "mpg"),cov = "cyl", data = mtcars)