2
$\begingroup$

Is the term “latent effect” in statistics has the same meaning as “latent variable”? If not, then what would be its substantive interpretation?

UPDATE:

There is a whole class of models called “Latent Gaussian models”, and that is where my confusion comes from. Basically, any article/book on Bayesian-hierarchical regression modeling mentions “latent effects”.
Below I have attached the picture of one article where this term is kind of introduced. The link to the whole article is here https://arxiv.org/pdf/1701.07844.pdf (hope it opens).

enter image description here

Improve this question
$\endgroup$
6
  • 1
    $\begingroup$ Can you include a citation or example of where the term latent effect has been used? $\endgroup$
    – Erik Ruzek
    Commented Sep 20, 2022 at 13:00
  • $\begingroup$ @ErikRuzek, updated the question $\endgroup$
    – Rustam
    Commented Sep 21, 2022 at 4:42
  • $\begingroup$ @ErikRuzek, any thoughts on it? $\endgroup$
    – Rustam
    Commented Sep 23, 2022 at 3:36
  • 1
    $\begingroup$ I believe these terms are referring to different things, but I am not familiar with the latent effects term whereas I am very familiar with latent variables. I am sorry. Given you are looking at papers using INLA, I would suggest posting this question to their discussion board (groups.google.com/g/r-inla-discussion-group). $\endgroup$
    – Erik Ruzek
    Commented Sep 24, 2022 at 21:07
  • 1
    $\begingroup$ I would only talk about a latent effect if I were considering a regression of one latent variable on another, for example in a structural equation model. But again, I'm not convinced they are using that term in the same way. $\endgroup$
    – Erik Ruzek
    Commented Oct 1, 2022 at 21:18

0

Reset to default

Your Answer

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