2
$\begingroup$

Can anyone tell me why and when should we use the "consec" contrast method of emmeans function (of emmeans package)? An exemple would help. thank you.

emmeans.out <- emmeans(model, consec ~ Time| Condition)
$\endgroup$
1
  • 1
    $\begingroup$ consec is provided as a contrast family that some users may find useful. It generally does not make sense unless the factor in question has an implied natural ordering of its levels. So actually he example you give could be a good one, if your main focus is comparing changes overtime increments. The usefulness of any method depends in great part on the scientific context. $\endgroup$
    – Russ Lenth
    Aug 5, 2021 at 15:20

1 Answer 1

2
$\begingroup$

I'll offer a succinct answer: it's useful when each level of a factor serves as a control for the next level.

For example, a drug experiment may have a factor dose, perhaps with levels 10, 20, 40, and 60, being the available ones from the drug manufacturer. Comparing successive doses answers questions about what happens if we increase the dose to the next level, and it seems reasonable to have those comparisons as a set of pre-planned contrasts.

Keep in mind that the more comparisons you test, the more conservative the tests become because of multiplicity corrections such as "mvt" or "tukey" (the latter being appropriate only when doing all pairwise comparisons). Any time you can reduce the number of tests without sacrificing potentially important findings, the better.

$\endgroup$

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.