Automatic essay scoring (AES) papers using AI often report spectacular results… and then the model is never used again, not even by its own authors. It's as if it was not so good after all.

A very simplified summary of my work.

So the plan was not to stop bots, but to turn bots into paid ads.

Caillou, no!!! In Grüter, T. (2006). Object (clitic) omission in L2 French: Mis-setting or missing surface inflection.

Research note out on parent's perception of the impact of COVID. Cute descriptive plots I must say. Also: hierarchical clustering instead of SEM = simpler and usually does the job. www.cjslpa.ca/detail.php?I...

promatch is a simple web app that does propensity score matching and outputs a .csv of your matched data, in wide or long format. #RCode gloignon.shinyapps.io/promatch/

New paper up, combining L2 French lexical bundle analysis and NLP with Randy Appel over at Waseda Tokyo, Beverly Baker from UOttawa and Angel Arias from Carleton U. www.sciencedirect.com/science/arti...

Grant panels in medicine and health:

😺We used linear regression to...
😡HOLD ON! dID yoU tESt 4 nORmaLiTy!?

😺We used AI to...
😍WOW! Sounds great. Say no more. No questions from me. I totally understand what's going on here.