Follow-Up Discusison with Guillaume Dumas

FPR CMB Network Meeting

Moderated by Dietrich Stout via Zoom

December 10, 2021


00:13:25 Brandon Kohrt: Jeff – this is a helpful recent critique of the standardized mean difference in treatment effects. That said, the alternatives proposed here don’t get at subgroup differences, etc. that you are raising:
00:14:02 Guillaume Dumas: Thanks Brandon!
00:15:13 Guillaume Dumas: We propose some guidelines in our recent paper on group-mean differences:
00:15:47 Jeffrey Snodgrass: @ty brandon
00:19:51 Jeffrey Snodgrass: @ty guillaume for your paper too, I’ll look at both of them
00:24:37 Daniel Lende: I really like Guillaume’s point about thinking about causality, and considering how to rework in a way that would expand clinical approaches. But also neuroscientific and anthropological appreciations of our causality models. What type of “interactive tool” would get at that interactive causality?
00:26:16 Sally Bernardina Seraphin: At present, our causality models are waaaay too simplistic.
00:26:45 Greg D: Interesting question, Daniel. I would love to hear if people are using different models or tools (or even software) to better map even theoretical models of multicausality.
00:28:34 Jeffrey Snodgrass: I’ve seen Eiko Fried taking an ecological approach to mental disorder
00:30:21 Jeffrey Snodgrass: I haven’t read this yet:
00:30:34 Jeffrey Snodgrass: Complexity Theory of Psychopathology, Olthof et al.
00:31:56 Jeffrey Snodgrass: Studying mental disorders as systems, not syndromes
00:32:42 Greg D: Thanks, Jeff. That looks great
00:32:42 Jeffrey Snodgrass: Here’s Fried’s actual piece:
00:41:29 Brandon Kohrt: Thanks for the PLoS paper, Guillaume. I was just going through that. It looks like you could also think about some of these shortcomings in group mean difference also in relation to challenges of cultural adaptation and validation of symptoms assessment tools.
00:42:07 Daniel Lende: Couple comments: (1) It might be interesting to think about quasi-experimental approaches, which would bring a predictive approach together with more social science approaches to recognizing that causality already comes packed into everything we do; (2) That “big data” in itself will have multiple dimensions, so a set of interviews with the same person to really figure out what has happened with that person is in itself a type of big data with high dimensionality.
00:43:34 Michael Lifshitz, Dr: Sorry I’m so late, I was having strange technical difficulties.
00:52:29 Greg D: Congratulations, Shinobu!
00:53:42 Jeffrey Snodgrass: @Brandon are you meaning applying measures developed in one group to another and the problems there?
00:56:00 Daniel Lende: For longitudinal approaches, I’ve long been interested in person-centered approaches (generally contrasted to variable centered approaches). The person bundles variables together, and that bundling is often discounted in many statistical approaches. David Magnusson was a long-term proponent of such approaches, which takes us away from debates about traits and subgroups and such. Here
00:56:21 Daniel Lende: Here is a pdf In Memoriam of Magnusson’s career.
01:04:26 Brandon Kohrt: Hi, All, I have to run. See you at the next meeting.
01:07:09 Laurence Kirmayer, Dr.: Here is some recent work form colleagues here” Neumann, A., Jolicoeur-Martineau, A., Szekely, E., Sallis, H. M., ODonnel, K., Greenwood, C. M., … & Tiemeier, H. (2020). Combined polygenic risk scores of different psychiatric traits predict general and specific psychopathology in childhood. medRxiv.
01:10:58 Constance A. Cummings: Power, R., Steinberg, S., Bjornsdottir, G. et al. Polygenic risk scores for schizophrenia and bipolar disorder predict creativity. Nat Neurosci 18, 953–955 (2015).
01:14:49 Guillaume Dumas: On PRS and ethnic diversity:
01:21:22 Shinobu Kitayama: I am sorry I have to go. The discussion was fascinating. Thank you so much, Guillaume!
01:25:33 Dietrich Stout (he/him): I’m afraid I’m going to have to make my way home now as well. Sorry to leave my “moderating” role early but clearly the discussion is moderating itself. Sorry to miss out on the rest!
01:32:34 Constance A. Cummings:
01:34:52 Constance A. Cummings:
01:34:54 Sally Bernardina Seraphin: Thank you @Marjorie. Do you have a reference I can follow up on to learn more?
01:35:22 Daniel Lende: I really like Jeff’s point that we need to start to compare across “meaningful groups”. Exciting! And then to look at individuals, activities, and more…
01:36:41 Greg D: I couldn’t agree more Laurence. To me it’s developmental dynamics not ethnic membership that is generalizable.
01:39:29 Greg D: Thanks, Guillaume!
01:39:47 Jeffrey Snodgrass: Merci Guillaume!
01:40:10 Sally Bernardina Seraphin: Thanks Guillaume!