[paper] finalize

[DominiqueMakowski]

No description provided.

[DominiqueMakowski]

We'll also need to list potential reviewers, any suggestions?

[strengejacke]

Do we want text tables:

or markdown tables?

[DominiqueMakowski]

mmh good question, markdown looks nicer but the former gives a more "accurate" representation of what people will get

[strengejacke]

• Vincent
• Jeffrey (jmgirard)
• TJ (Mahr)
• profandyfield

wasn't there a reviewer list on JOSS, or have they dismissed it?

[DominiqueMakowski]

Maybe let's keep text tables

[strengejacke]

Maybe let's keep text tables

Agreed.

Regarding your comment about generalization of results (TODO: is this correct?): basically, G-computation, like propensity scores, attempts at conducting a "pseudo-randomization" for observational data, where randomization is not possible. This is why we have a "hypothetical", broader population that is not equal to the sample. It should reduce confounder bias between groups (levels of focal predictors), which means, it's more "representative" of a population, i.e. more generalizable.

[strengejacke]

@mattansb @bwiernik @IndrajeetPatil @rempsyc You're listed as co-authors, would you approve the paper or would you like to revise anything?

[DominiqueMakowski]

one rule: do not talk about the marginalization option names here 😁

[strengejacke]

one rule: do not talk about the marginalization option names here 😁

_{Dom thinking: just approve your affiliation, that's enough...}

[mattansb]

LGTM

[DominiqueMakowski]

Should we add a paragraph about estimate_grouplevel() and like BLUPs vs. random coefs, this would also be nice to have some clear info about that for people

[bwiernik]

I'll give it a read when I'm flying home Saturday

[strengejacke]

Should we add a paragraph about estimate_grouplevel() and like BLUPs vs. random coefs, this would also be nice to have some clear info about that for people

No strong opinion here, but I think the paper is already quite long for a JOSS paper. I would probably put more effort into another paper that could be submitted to a (listed) journal.

[DominiqueMakowski]

I would probably put more effort into another paper that could be submitted to a (listed) journal

Mmh fair fair I'll see if I can throw in 2 sentences about that for the sake of comprehensiveness but we can keep our efforts for another paper

[bwiernik]

Sorry Dom, but I think we should nail down the marginalization names before submitting. Which issue has the discussion on that again?

[strengejacke]

Here
#400

[strengejacke]

What's the current state of the paper draft? Still need for discussing options/names or parts of the paper itself?

[DominiqueMakowski]

afaic it's good to be submitted

[DominiqueMakowski]

Since I've made some changes to the docstrings of grouplevel, I'm going to merge that @strengejacke ? so that we can merge it in the CRAN update

[rempsyc]

would you approve

approve (edit: also waiting on @DominiqueMakowski 's approval for #227 hihi...)

[strengejacke]

Since I've made some changes to the docstrings of grouplevel, I'm going to merge that @strengejacke ? so that we can merge it in the CRAN update

Sounds good

[strengejacke]

Since I've made some changes to the docstrings of grouplevel, I'm going to merge that @strengejacke ? so that we can merge it in the CRAN update

Can you check the docstrings again? Due the many merging of main into this branch, I hope they're still up to date (and the docstrings were also partly copied to the submission branch)