1
$\begingroup$

When I get 2 mixed linear Models for a comparison between them

For example, (B, C, D are factors)

Mdoel 1 <- lmer(A ~ B * C * D + (1|individual), data = Data_1) in script Outcome in console: A ~ B * C * D + (1 | individual)

Model 2 <- update(Model 1, . ~ . - B : C : D) in script Outcome in console : A ~ B + C + D + (1 | individual) + B:C + B:D + C:D

  1. Also, B,C,D are for fixed effects and (1|individual) is for a random effect

What is the difference between A ~ B + C + D + (1|individual) and A ~ BCD + (1|individual)?

  1. By Model 2 what do mean ". ~ ." and "- B : C : D" both in Script?

2-2. A ~ B + C + D + (1 | individual) + B:C + B:D + C:D What does mean " B:C + B:D + C:D "?

$\endgroup$

1 Answer 1

1
$\begingroup$

What is the difference between A ~ B + C + D + (1|individual) and A ~ BCD + (1|individual)?

I think you mean B * C * D, not BCD. BCD would represent a variable called BCD, wheras B * C * D means "fit fixed effects for B, C and D, and all the interactions between them" (that is, the three 2-way interactions plus the 3 way interaction). So the difference betwwen A ~ B + C + D and A ~ B * C * D is that in the former we fit fixed effects for the each of the variables while in the latter we also fit all the interactions.

By Model2 what do mean ". ~ ." and "- B : C : D" both in Script?

You are telling the software to update model 1 by removing the 3-way interaction B:C:D(because you used the minus symbol -), so that model 2 contains only the 2-way interactions (and the main effects)

What does mean "B:C + B:D + C:D"?

B:C + B:D + C:D tells the software to fit the three 2-way interactions between the variables. Note that

~ B * C * D

is exactly the same as:

~ B + C + D + B:C + B:D + C:D
$\endgroup$
2
  • $\begingroup$ Super super thank you I did understand totally :) I have one more question so I did AIC and BIC for model comparison model 1 has more information than model 2 cuz model 2 has no information about 3-way interaction but the outcome of AIC and BIC tells the model 2 is better is it possible? $\endgroup$
    – Taede17
    Jul 10, 2021 at 21:28
  • $\begingroup$ You're welcome :) Yes that is possible ! $\endgroup$ Jul 11, 2021 at 9:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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