I just wanted to make sure I'm interpreting my lme() output correctly. I've set up a null model because I was told I could find the within-persons SD this way (Nelzek, 2012 MLM chapter). But the chapter is written in formulas, not in R code.

Here's the model I used; subjectID is the grouping variable for person:

nullmodel <- lme(dv1 ~ 1, data = dataset, random = ~ 1 | subjectID,  
                 na.action = na.exclude, method = "ML")

And here's the output of the summary() method:

Linear mixed-effects model fit by maximum likelihood
Data: dataset 
AIC      BIC    logLik
34404.33 34423.05 -17199.16

Random effects:
Formula: ~1 | subjectID
          (Intercept) Residual
StdDev:    13.39356   21.844

Fixed effects: dv1 ~ 1 
               Value Std.Error   DF  t-value p-value
(Intercept) 37.54255  1.336242 3677 28.09562       0

Standardized Within-Group Residuals:
       Min         Q1        Med         Q3        Max 
-2.7019919 -0.6909552 -0.1045376  0.6288470  3.4520256

I assume the within-person SD for the dependent variable (dv1) is the (Intercept) StdDev, but just wanted to check. Or if I'm interpreting this wrong?


The standard deviation for the random intercepts that equals 13.39356 quantifies the variability between subjects.

The residuals standard deviation (also known as the residual standard error) that equals 21.844 quantifies the variability of the outcome within a subject.

  • $\begingroup$ Thanks! I got the same answer on stack and just wanted to confirm :) $\endgroup$ – alessothegreat Sep 26 at 19:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.