statistically adjustment in R using multiple linear regression-CodePudding

I have a question regarding statistically adjustment.

I have a multiple linear regression modell like this

bodysize <- data.frame(
  stringsAsFactors = FALSE,
  humansize = c(166,159,161,162,155,155,178,165,163,169,170,171,172,172,173,166,167,157,159,160,178,173,168,167,166,165,166,164,162,163, 190,191,192,193,195,199,189,188,187,185,183,181,188,189,200,201,198,179,178,177,176,177,173,175,176,188,185,183,181,196),
  Temperature = c(25,23,24,25,24,23,12,19,20,15,13,9,9,8,12,11,26,29,30,11,12,13,10,14,15,16,17,13,11,9,7,8,9,9,9,10,12,13,12,14,13,25,13,7,6,8,7,6,23,24,25,26,27,27,23,21,12,13,14,10),
airpollution = c(95,93,84,85,88,93,52,39,72,58,27,38,34,23,62,61,56,99,99,44,42,43,30,44,45,76,67,63,51,29,27,28,39,29,29,30,32,33,32,54,53,85,63,17,16,18,17,16,76,84,85,87,67,67,63,81,32,13,34,40),
SES= c(25,13,34,25,28,13,32,79,12,68,87,58,54,53,12,11,36,9,9,74,72,43,30,44,45,16,17,53,51,89,87,78,69,79,89,60,52,33,32,54,53,65,63,67,66,78,77,67,20,4,5,7,7,7,13,11,62,73,34,40),
Gender = c("female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","female","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male","male")
)


humansize <- lm(humansize~Temperature airpollution SES factor(Gender), data = bodysize)
summary(humansize)
confint(humansize)

This is my modell. I would do a statistical adjustment without the variables temperature and airpollution, because they are confounders. How can I do that?

I would like to know how human size would be if temperature and air pollution are adjusted for

CodePudding user response：

Typically 'statistical adjusting' for some covariates means running a linear model with said covariates and then taking the residuals from that model. In your case, it would simply be

humansize$residuals

  6.43345246  -1.64591706   0.49132162   2.36454718  -5.34769995  -5.64591706 
           7            8            9           10           11           12 
  9.32077018  -0.15042732   0.23298624   1.56024709   0.59173459  -0.35246626 
          13           14           15           16           17           18 
  0.70891584  -0.01085447   4.83439654  -2.81646849   7.58633582   0.48163364 
          19           20           21           22           23           24 
  3.14784419 -10.33447884   8.36242277   4.68036940  -2.11877044  -0.66876558 
          25           26           27           28           29           30 
 -1.01790055  -0.49323807   1.08872167  -4.40418057  -7.77481591  -9.10379867 
          31           32           33           34           35           36 
 -2.36715056  -0.49392500   1.44820585   2.15694003   3.93457950   9.25252613 
          37           38           39           40           41           42 
  0.77661671   0.87220281  -0.77866222  -1.78384266  -4.43470769   1.51348321 
          43           44           45           46           47           48 
  0.41183706  -2.99133477   7.35780020   9.43716972   5.78630469 -13.66443585 
          49           50           51           52           53           54 
 -1.88033025  -1.80321862  -2.15235359  -0.51683409  -3.98843410  -1.98843410 
          55           56           57           58           59           60 
 -3.81425475   7.02182576  -3.44574382  -5.15504989  -5.47693216   6.76615248