#DataMethods #OpenScience
#GESISMethodsHub: An open portal for computational social science research—find tools, tutorials, and code, directly usable in your browser. Methods Hub makes it easy to learn, apply, and reproduce computational methods in the Social Sciences.
methodshub.gesis.org/
#GESISMethodsHub: An open portal for computational social science research—find tools, tutorials, and code, directly usable in your browser. Methods Hub makes it easy to learn, apply, and reproduce computational methods in the Social Sciences.
methodshub.gesis.org/
October 15, 2025 at 8:00 AM
#DataMethods #OpenScience
#GESISMethodsHub: An open portal for computational social science research—find tools, tutorials, and code, directly usable in your browser. Methods Hub makes it easy to learn, apply, and reproduce computational methods in the Social Sciences.
methodshub.gesis.org/
#GESISMethodsHub: An open portal for computational social science research—find tools, tutorials, and code, directly usable in your browser. Methods Hub makes it easy to learn, apply, and reproduce computational methods in the Social Sciences.
methodshub.gesis.org/
He has a datamethods forum where it looks like its been mentioned quite a few times, discourse.datamethods.org/search?q=par.... He's also on bluesky and pretty responsive to questions.
Search results for 'partial proportional odds' - Datamethods Discussion Forum
This is a place for discussions and Q&A about data-related issues and quantitative methods including study design, data analysis, and interpretation.
discourse.datamethods.org
September 26, 2025 at 1:54 PM
He has a datamethods forum where it looks like its been mentioned quite a few times, discourse.datamethods.org/search?q=par.... He's also on bluesky and pretty responsive to questions.
Also check out the datamethods discussion: discourse.datamethods.org/t/leo-breima...
Leo Breiman, the Rashomon Effect, and the Occam Dilemma
Cynthia Rudin, who works in making interpretable machine learning models has this new paper about Leo Breiman’s famous two cultures paper. I thought it was insightful, although I am not sure I underst...
discourse.datamethods.org
August 7, 2025 at 2:58 PM
Also check out the datamethods discussion: discourse.datamethods.org/t/leo-breima...
I learned of a new type of effect the other day: Rashomon effect (i.e., the higher the diversity of models that can near equally well explain the same data) then the more likely it becomes that simple models will be included among them. #datamethods #statistics #statssky
August 7, 2025 at 2:57 PM
I learned of a new type of effect the other day: Rashomon effect (i.e., the higher the diversity of models that can near equally well explain the same data) then the more likely it becomes that simple models will be included among them. #datamethods #statistics #statssky
New discussion topic on weighted analyses on datamethods. Join the controversy there. discourse.datamethods.org/t/illogic-of...
#statistics #StatsSky #EpiSky
#statistics #StatsSky #EpiSky
Illogic of Weighting While Estimating Model Parameters
Consider a situation where some types of subjects are over- or under-sampled from a population where the sampling probabilities are known. For example, to increase efficiency we might sample a greate...
discourse.datamethods.org
July 29, 2025 at 12:16 PM
New discussion topic on weighted analyses on datamethods. Join the controversy there. discourse.datamethods.org/t/illogic-of...
#statistics #StatsSky #EpiSky
#statistics #StatsSky #EpiSky
decided to tell all of you that I have a problem and it is forgetting that i do not have three brains and nine arms. i sat down to do some regression btw, thats why i went to datamethods because there was some thing about sample selection that i read last week i wanted to review. this has been […]
Original post on sigmoid.social
sigmoid.social
June 14, 2025 at 9:22 AM
decided to tell all of you that I have a problem and it is forgetting that i do not have three brains and nine arms. i sat down to do some regression btw, thats why i went to datamethods because there was some thing about sample selection that i read last week i wanted to review. this has been […]
got something delivered ordered food started reading a datamethods thread about stats and AI jumped to a paper that was linked, jumped back to a textbook that was linked, wrote up a project brief, read 30% of that article that id opened, checked mastodon, went back to the article and did some […]
Original post on sigmoid.social
sigmoid.social
June 14, 2025 at 9:13 AM
got something delivered ordered food started reading a datamethods thread about stats and AI jumped to a paper that was linked, jumped back to a textbook that was linked, wrote up a project brief, read 30% of that article that id opened, checked mastodon, went back to the article and did some […]
I asked Grok to “Summarize REMAP CAP using two DAGs”
But Grok only looks at work I did on X and discussions at datamethods and provides a
Well organized summary. Grok is a sophisticated text review parrot. Which is nice.
discourse.datamethods.org/t/the-petty-...
But Grok only looks at work I did on X and discussions at datamethods and provides a
Well organized summary. Grok is a sophisticated text review parrot. Which is nice.
discourse.datamethods.org/t/the-petty-...
The Petty/Bone RCT
Looking forward and seeking a solution hypothesis to the present critical care research crisis highlighted by the recent REMAP CAP, what are the thoughts of adding symbolic causal modeling with DAGs o...
discourse.datamethods.org
June 3, 2025 at 11:48 AM
I asked Grok to “Summarize REMAP CAP using two DAGs”
But Grok only looks at work I did on X and discussions at datamethods and provides a
Well organized summary. Grok is a sophisticated text review parrot. Which is nice.
discourse.datamethods.org/t/the-petty-...
But Grok only looks at work I did on X and discussions at datamethods and provides a
Well organized summary. Grok is a sophisticated text review parrot. Which is nice.
discourse.datamethods.org/t/the-petty-...
The DataMethods forum is a great resource for technical questions like yours.
discourse.datamethods.org/latest
a snippet:
discourse.datamethods.org/latest
a snippet:
June 2, 2025 at 10:07 AM
The DataMethods forum is a great resource for technical questions like yours.
discourse.datamethods.org/latest
a snippet:
discourse.datamethods.org/latest
a snippet:
No one will discuss or retweet PettyBone RCT mimics. But your generation needs the courage to break out. Study and teach Pearl’s symbolic causal modeling
Send your pul fellows to see this alternative view explaining REMAP CAP at datamethods. 6.8k views
discourse.datamethods.org/t/the-petty-...
Send your pul fellows to see this alternative view explaining REMAP CAP at datamethods. 6.8k views
discourse.datamethods.org/t/the-petty-...
The Petty/Bone RCT
Looking forward and seeking a solution hypothesis to the present critical care research crisis highlighted by the recent REMAP CAP, what are the thoughts of adding symbolic causal modeling with DAGs o...
discourse.datamethods.org
May 29, 2025 at 2:12 PM
No one will discuss or retweet PettyBone RCT mimics. But your generation needs the courage to break out. Study and teach Pearl’s symbolic causal modeling
Send your pul fellows to see this alternative view explaining REMAP CAP at datamethods. 6.8k views
discourse.datamethods.org/t/the-petty-...
Send your pul fellows to see this alternative view explaining REMAP CAP at datamethods. 6.8k views
discourse.datamethods.org/t/the-petty-...
Or bring your excellent points to Vanderbilt Datamethods forum.
discourse.datamethods.org/t/the-petty-...
discourse.datamethods.org/t/the-petty-...
The Petty/Bone RCT
Yes. The Petty/Bone RCT is a term used for a unique pathological modification of the RCT which dominates critical care science. Two USA pulmonologists, Thomas Petty (1960s) and Roger Bone (1980s) de...
discourse.datamethods.org
May 27, 2025 at 8:12 PM
Or bring your excellent points to Vanderbilt Datamethods forum.
discourse.datamethods.org/t/the-petty-...
discourse.datamethods.org/t/the-petty-...
After two years of trying to avoid this discussion, I just necroed *that thread* on datamethods (discourse.datamethods.org/t/should-one...) in order to share an excellent preprint by philosopher Veli-Pekka Parkkinen (philsci-archive.pitt.edu/24785/1/efme...)
Should one derive risk difference from the odds ratio?
I urge all readers of this thread to read the excellent new preprint from philosopher of science Veli-Pekka Parkkinen, “Choice of effect measure, extrapolation, and decision-making in patient care an...
discourse.datamethods.org
February 26, 2025 at 10:00 AM
After two years of trying to avoid this discussion, I just necroed *that thread* on datamethods (discourse.datamethods.org/t/should-one...) in order to share an excellent preprint by philosopher Veli-Pekka Parkkinen (philsci-archive.pitt.edu/24785/1/efme...)
Ehud, thanks for responding to an older post—I also saw your comment on Datamethods. It seems to me that as long as one conducts an in-depth examination of the propensity score distribution (even when using matching), the method can still be valid and useful.
February 13, 2025 at 3:09 PM
Ehud, thanks for responding to an older post—I also saw your comment on Datamethods. It seems to me that as long as one conducts an in-depth examination of the propensity score distribution (even when using matching), the method can still be valid and useful.
First, note Andrew's post discusses propensity score matching while Solomon uses covariate-distance-based matching.
but more generally, Frank (and by extension Datamethods) quite opposes matching/weighting-based techniques. His argument is about data inefficiency and inflated standard errors,
but more generally, Frank (and by extension Datamethods) quite opposes matching/weighting-based techniques. His argument is about data inefficiency and inflated standard errors,
February 13, 2025 at 9:02 AM
First, note Andrew's post discusses propensity score matching while Solomon uses covariate-distance-based matching.
but more generally, Frank (and by extension Datamethods) quite opposes matching/weighting-based techniques. His argument is about data inefficiency and inflated standard errors,
but more generally, Frank (and by extension Datamethods) quite opposes matching/weighting-based techniques. His argument is about data inefficiency and inflated standard errors,
HT @f2harrell.bsky.social for good discussion at Datamethods
February 9, 2025 at 2:56 PM
HT @f2harrell.bsky.social for good discussion at Datamethods
I consider Dr. Palmer a hero for presenting such a well reasoned debate at datamethods. His points were much needed and have been very well received.
It takes courage to debate the fundamental methodology. Almost no one will. It would be great if more thought leaders enter the discussion.
It takes courage to debate the fundamental methodology. Almost no one will. It would be great if more thought leaders enter the discussion.
January 28, 2025 at 3:39 PM
I consider Dr. Palmer a hero for presenting such a well reasoned debate at datamethods. His points were much needed and have been very well received.
It takes courage to debate the fundamental methodology. Almost no one will. It would be great if more thought leaders enter the discussion.
It takes courage to debate the fundamental methodology. Almost no one will. It would be great if more thought leaders enter the discussion.
This has been quite a lively debate on trials in intensive care. I don't think the "pathological science" chat is productive, but I'm always up for a healthy scientific debate. Thoughts on datamethods: discourse.datamethods.org/t/the-petty-... @statsepi.bsky.social @load-dependent.bsky.social
January 26, 2025 at 4:20 PM
This has been quite a lively debate on trials in intensive care. I don't think the "pathological science" chat is productive, but I'm always up for a healthy scientific debate. Thoughts on datamethods: discourse.datamethods.org/t/the-petty-... @statsepi.bsky.social @load-dependent.bsky.social
But thanks for your thoughts. Probably good to end the discussion unless you have a comments about CCM science.
Come over to datamethods. There is no provocation there. It’s not needed. It’s all science.
Come over to datamethods. There is no provocation there. It’s not needed. It’s all science.
January 25, 2025 at 9:46 PM
But thanks for your thoughts. Probably good to end the discussion unless you have a comments about CCM science.
Come over to datamethods. There is no provocation there. It’s not needed. It’s all science.
Come over to datamethods. There is no provocation there. It’s not needed. It’s all science.
Your point is discussed at datamethods. Please bring your excellent thoughts there.
The mRCT are generally negative or reversed also and this is predicted in PettyBone RCT due to greater dilution of a fortuitously responsive disease in the disease mix captured by the triage thresholds in the mRCT
The mRCT are generally negative or reversed also and this is predicted in PettyBone RCT due to greater dilution of a fortuitously responsive disease in the disease mix captured by the triage thresholds in the mRCT
January 23, 2025 at 1:15 PM
Your point is discussed at datamethods. Please bring your excellent thoughts there.
The mRCT are generally negative or reversed also and this is predicted in PettyBone RCT due to greater dilution of a fortuitously responsive disease in the disease mix captured by the triage thresholds in the mRCT
The mRCT are generally negative or reversed also and this is predicted in PettyBone RCT due to greater dilution of a fortuitously responsive disease in the disease mix captured by the triage thresholds in the mRCT
I was thinking about the best way to mitigate bias when bootstraping confidence intervals (before or after imputation) over the weekend and found this excellent post on #datamethods #statssky
Extremely thorough response by @Isabella_Ghement
discourse.datamethods.org/t/bootstrapp...
Extremely thorough response by @Isabella_Ghement
discourse.datamethods.org/t/bootstrapp...
Bootstrapping CIs in multiply imputed data
Background: I plan to estimate the average causal effect of a continuous exposure on my outcome using the parametric g-formula (i.e., standardized mean difference), and I need to bootstrap the 95% CIs...
discourse.datamethods.org
January 13, 2025 at 5:17 PM
I was thinking about the best way to mitigate bias when bootstraping confidence intervals (before or after imputation) over the weekend and found this excellent post on #datamethods #statssky
Extremely thorough response by @Isabella_Ghement
discourse.datamethods.org/t/bootstrapp...
Extremely thorough response by @Isabella_Ghement
discourse.datamethods.org/t/bootstrapp...
Great, so pls send colleagues to Datamethods to discuss with statisticians. Why wait? Let’s end the PettyBone paradigm
ARDS & sepsis will remain as cognitive constructs for medED & severity measures
Common Pathways will be discovered by starting with disease & adding, not starting with Delphi.
ARDS & sepsis will remain as cognitive constructs for medED & severity measures
Common Pathways will be discovered by starting with disease & adding, not starting with Delphi.
January 7, 2025 at 7:39 PM
Great, so pls send colleagues to Datamethods to discuss with statisticians. Why wait? Let’s end the PettyBone paradigm
ARDS & sepsis will remain as cognitive constructs for medED & severity measures
Common Pathways will be discovered by starting with disease & adding, not starting with Delphi.
ARDS & sepsis will remain as cognitive constructs for medED & severity measures
Common Pathways will be discovered by starting with disease & adding, not starting with Delphi.
This is your time. Petty & Bone’s ideas were from their generation.
Start with real inquiry. Send your colleagues to datamethods to discuss & debate. Those are world class statisticians who are not in-the-box at datamethods
Science only self-corrects if the youth correct their mentors mistakes
Start with real inquiry. Send your colleagues to datamethods to discuss & debate. Those are world class statisticians who are not in-the-box at datamethods
Science only self-corrects if the youth correct their mentors mistakes
January 7, 2025 at 2:48 PM
This is your time. Petty & Bone’s ideas were from their generation.
Start with real inquiry. Send your colleagues to datamethods to discuss & debate. Those are world class statisticians who are not in-the-box at datamethods
Science only self-corrects if the youth correct their mentors mistakes
Start with real inquiry. Send your colleagues to datamethods to discuss & debate. Those are world class statisticians who are not in-the-box at datamethods
Science only self-corrects if the youth correct their mentors mistakes
Thanks for coming to datamethods.
See if you can find colleagues or mentor who will publicly defend the PettyBoneRCT. I have not found anyone yet. However if you do, please send them as we would welcome their input. In the comments I will put the link to send them.
See if you can find colleagues or mentor who will publicly defend the PettyBoneRCT. I have not found anyone yet. However if you do, please send them as we would welcome their input. In the comments I will put the link to send them.
January 7, 2025 at 4:06 AM
Thanks for coming to datamethods.
See if you can find colleagues or mentor who will publicly defend the PettyBoneRCT. I have not found anyone yet. However if you do, please send them as we would welcome their input. In the comments I will put the link to send them.
See if you can find colleagues or mentor who will publicly defend the PettyBoneRCT. I have not found anyone yet. However if you do, please send them as we would welcome their input. In the comments I will put the link to send them.