Brady West
@bradytwest.bsky.social
Research Professor @ Univ. of MI Institute for Social Research (ISR) and Department of Biostatistics. Usually discussing statistics, surveys, sports, music, and leadership. Personal website: www.umich.edu/~bwest
This was an absolute all-timer of a Gus Johnson call, which is appropriate given how utterly incredible the catch was. Crazy!
November 8, 2025 at 8:44 PM
This was an absolute all-timer of a Gus Johnson call, which is appropriate given how utterly incredible the catch was. Crazy!
Reposted by Brady West
Reposted by Brady West
LLMs are now widely used in social science as stand-ins for humans—assuming they can produce realistic, human-like text
But... can they? We don’t actually know.
In our new study, we develop a Computational Turing Test.
And our findings are striking:
LLMs may be far less human-like than we think.🧵
But... can they? We don’t actually know.
In our new study, we develop a Computational Turing Test.
And our findings are striking:
LLMs may be far less human-like than we think.🧵
Computational Turing Test Reveals Systematic Differences Between Human and AI Language
Large language models (LLMs) are increasingly used in the social sciences to simulate human behavior, based on the assumption that they can generate realistic, human-like text. Yet this assumption rem...
arxiv.org
November 7, 2025 at 11:13 AM
LLMs are now widely used in social science as stand-ins for humans—assuming they can produce realistic, human-like text
But... can they? We don’t actually know.
In our new study, we develop a Computational Turing Test.
And our findings are striking:
LLMs may be far less human-like than we think.🧵
But... can they? We don’t actually know.
In our new study, we develop a Computational Turing Test.
And our findings are striking:
LLMs may be far less human-like than we think.🧵
Reposted by Brady West
Struggling with missing data? Join Paul Allison for "Missing Data Using R", a 4-week on-demand seminar starting Nov. 17. You'll learn maximum likelihood & multiple imputation—two of the most powerful #missingdata solutions using #Rstats.
Missing Data Using R | On-Demand Seminar | Statistical Horizons
Paul Allison, Ph.D., teaches this self-paced course on missing data using R, covering multiple imputation and maximum likelihood for accurate, unbiased results.
statisticalhorizons.com
November 6, 2025 at 2:47 PM
Struggling with missing data? Join Paul Allison for "Missing Data Using R", a 4-week on-demand seminar starting Nov. 17. You'll learn maximum likelihood & multiple imputation—two of the most powerful #missingdata solutions using #Rstats.
Reposted by Brady West
New blog post: open-source software packages have surprising problems with the way they calculate weighted medians and other quantiles.
www.practicalsignificance.com/posts/weight...
#rstats #julialang
www.practicalsignificance.com/posts/weight...
#rstats #julialang
Weighted Quantile Weirdness and Bugs – Practical Significance
Computing quantiles is surprisingly complicated. It gets much weirder when you use weights, and popular software behaves in surprising ways that might trouble you.
www.practicalsignificance.com
November 5, 2025 at 4:30 PM
New blog post: open-source software packages have surprising problems with the way they calculate weighted medians and other quantiles.
www.practicalsignificance.com/posts/weight...
#rstats #julialang
www.practicalsignificance.com/posts/weight...
#rstats #julialang
Reposted by Brady West
Data scientists perform last rites for ‘dearly departed datasets’ in 2nd Trump administration
Thank you @mikeysid.bsky.social for covering the loss of public data.
apnews.com/article/cens...
Thank you @mikeysid.bsky.social for covering the loss of public data.
apnews.com/article/cens...
Data scientists perform last rites for 'dearly departed datasets' in 2nd Trump administration
A group of U.S. data scientists has published a list of federal datasets that have been altered or removed since President Donald Trump returned to the White House.
apnews.com
November 5, 2025 at 6:04 PM
Data scientists perform last rites for ‘dearly departed datasets’ in 2nd Trump administration
Thank you @mikeysid.bsky.social for covering the loss of public data.
apnews.com/article/cens...
Thank you @mikeysid.bsky.social for covering the loss of public data.
apnews.com/article/cens...
Sooo Michigan Hoops might be as good as advertised this year? 😳
November 4, 2025 at 2:26 AM
Sooo Michigan Hoops might be as good as advertised this year? 😳
Reposted by Brady West
@tschepis.bsky.social, Philip Veliz, @bradytwest.bsky.social, Jason A. Ford, and Sean Esteban McCabe used national survey data to identify motive classes for nonmedical tranquilizer use among US teens & young adults and explored links to other substance use and mental health outcomes myumi.ch/XykGZ
Sage Journals: Discover world-class research
Subscription and open access journals from Sage, the world's leading independent academic publisher.
myumi.ch
November 3, 2025 at 9:17 PM
@tschepis.bsky.social, Philip Veliz, @bradytwest.bsky.social, Jason A. Ford, and Sean Esteban McCabe used national survey data to identify motive classes for nonmedical tranquilizer use among US teens & young adults and explored links to other substance use and mental health outcomes myumi.ch/XykGZ
Reposted by Brady West
This, by @vincentab.bsky.social, is the most straight-up *useful* book I’ve bought in a long while.
November 3, 2025 at 6:40 PM
This, by @vincentab.bsky.social, is the most straight-up *useful* book I’ve bought in a long while.
Reposted by Brady West
Roderick Little contributed Pattern-Mixture Models for Missing Data to the JAMA Guide to Statistics and Methods. In the essay, Little discusses the advantages and limitations of pattern-mixture models which allow the predictive distribution of missing variables myumi.ch/bVwZk
Pattern-Mixture Models for Missing Data
This JAMA Guide to Statistics and Methods article discusses designing and implementing clinical trials to minimize the number of missing data.
myumi.ch
November 3, 2025 at 3:55 PM
Roderick Little contributed Pattern-Mixture Models for Missing Data to the JAMA Guide to Statistics and Methods. In the essay, Little discusses the advantages and limitations of pattern-mixture models which allow the predictive distribution of missing variables myumi.ch/bVwZk
Reposted by Brady West
Mark your calendar for next week's free, online half-day PDHP Workshop with @alussier.bsky.social: demonstrating the Structured Life Course Modelling Approach (SLCMA) applied to High Dimensional Data. Register at psc.isr.umich.edu/events/pdhp-...
November 3, 2025 at 3:48 PM
Mark your calendar for next week's free, online half-day PDHP Workshop with @alussier.bsky.social: demonstrating the Structured Life Course Modelling Approach (SLCMA) applied to High Dimensional Data. Register at psc.isr.umich.edu/events/pdhp-...
This has been a very disappointing meltdown by all three phases for the Lions. Just a lot of bad football this weekend. 👎 But nice to see JJ have a solid game.
November 2, 2025 at 8:56 PM
This has been a very disappointing meltdown by all three phases for the Lions. Just a lot of bad football this weekend. 👎 But nice to see JJ have a solid game.
This is big if true…
November 2, 2025 at 1:45 PM
This is big if true…
Reposted by Brady West
The target population for the research question is the population of people who would satisfy the eligibility criteria. “
Wrong x 2
1) The eligibility criteria define who doesn’t come into the trial not who does.
2) We generalise from quite different trials to populations eg bioequivalence trials.
Wrong x 2
1) The eligibility criteria define who doesn’t come into the trial not who does.
2) We generalise from quite different trials to populations eg bioequivalence trials.
November 1, 2025 at 7:03 PM
The target population for the research question is the population of people who would satisfy the eligibility criteria. “
Wrong x 2
1) The eligibility criteria define who doesn’t come into the trial not who does.
2) We generalise from quite different trials to populations eg bioequivalence trials.
Wrong x 2
1) The eligibility criteria define who doesn’t come into the trial not who does.
2) We generalise from quite different trials to populations eg bioequivalence trials.
I would encourage you to at least read the Executive Summary of this excellent report. 💯
🚨It's finally here!🚨
AAPOR's Taskforce on 2024 Pre-Election Polling report is out!
Full report: /https://aapor.org/wp-content/uploads/2025/10/AAPOR-Task-Force-on-2024-Pre-Election-Polling_Report.pdf
Executive summary: aapor.org/wp-content/u...
AAPOR's Taskforce on 2024 Pre-Election Polling report is out!
Full report: /https://aapor.org/wp-content/uploads/2025/10/AAPOR-Task-Force-on-2024-Pre-Election-Polling_Report.pdf
Executive summary: aapor.org/wp-content/u...
aapor.org
October 29, 2025 at 7:57 PM
I would encourage you to at least read the Executive Summary of this excellent report. 💯
Reposted by Brady West
🎉 @rpsychologist.com 's PowerLMM.js is the online statistics application of the year 2025 🎉
powerlmmjs.rpsychologist.com
- Calculate power (etc) for multilevel models
- Examine effects of dropout and other important parameters
- Fast! (Instant results)
powerlmmjs.rpsychologist.com
- Calculate power (etc) for multilevel models
- Examine effects of dropout and other important parameters
- Fast! (Instant results)
October 28, 2025 at 2:37 PM
🎉 @rpsychologist.com 's PowerLMM.js is the online statistics application of the year 2025 🎉
powerlmmjs.rpsychologist.com
- Calculate power (etc) for multilevel models
- Examine effects of dropout and other important parameters
- Fast! (Instant results)
powerlmmjs.rpsychologist.com
- Calculate power (etc) for multilevel models
- Examine effects of dropout and other important parameters
- Fast! (Instant results)
Reposted by Brady West
Ever wondered how to get better survey feedback from respondents? 🧐
👉 Check out our new @jssam.bsky.social paper together with @jkhoehne.bsky.social @jessicakuhlm.bsky.social testing different (1) visual designs and (2) answer formats of #FinalCommentQuestions.
🌐 academic.oup.com/jssam/advanc...
👉 Check out our new @jssam.bsky.social paper together with @jkhoehne.bsky.social @jessicakuhlm.bsky.social testing different (1) visual designs and (2) answer formats of #FinalCommentQuestions.
🌐 academic.oup.com/jssam/advanc...
Asking for Feedback: Innovating Final Comment Questions in Self-Administered Web Surveys
Abstract. Web surveys frequently include so-called “final comment questions” (FCQs) to provide respondents the opportunity to express their experiences wit
academic.oup.com
October 29, 2025 at 7:38 AM
Ever wondered how to get better survey feedback from respondents? 🧐
👉 Check out our new @jssam.bsky.social paper together with @jkhoehne.bsky.social @jessicakuhlm.bsky.social testing different (1) visual designs and (2) answer formats of #FinalCommentQuestions.
🌐 academic.oup.com/jssam/advanc...
👉 Check out our new @jssam.bsky.social paper together with @jkhoehne.bsky.social @jessicakuhlm.bsky.social testing different (1) visual designs and (2) answer formats of #FinalCommentQuestions.
🌐 academic.oup.com/jssam/advanc...
More important computational science that may not happen because it mentions diversity and inclusion. Fun times…and yes, definitely worth a donation if you are a Python user.
The Python Software Foundation got a competitive US research grant, but it came with a condition that they recant and abjure any diversity and inclusion ideas, on penalty of having to repay the money.
Obviously this is not desirable or safe, so no grant.
Donations would help them not regret this
Obviously this is not desirable or safe, so no grant.
Donations would help them not regret this
October 28, 2025 at 10:09 AM
More important computational science that may not happen because it mentions diversity and inclusion. Fun times…and yes, definitely worth a donation if you are a Python user.
Reposted by Brady West
📘 𝗡𝗲𝘄 𝗦𝗽𝗲𝗰𝗶𝗮𝗹 𝗜𝘀𝘀𝘂𝗲 𝗼𝗳 𝙎𝙪𝙧𝙫𝙚𝙮 𝙍𝙚𝙨𝙚𝙖𝙧𝙘𝙝 𝙈𝙚𝙩𝙝𝙤𝙙𝙨:
“Survey Climate and Trust in Scientific Surveys.”
Explore how the relationship between surveys and society has evolved.
🔗 Read the 𝗳𝘂𝗹𝗹 𝗶𝘀𝘀𝘂𝗲 here:
ojs.ub.uni-konstanz.de/srm/issue/vi...
#SurveyResearch #SurveyMethods #SurveyClimate #ESRA
“Survey Climate and Trust in Scientific Surveys.”
Explore how the relationship between surveys and society has evolved.
🔗 Read the 𝗳𝘂𝗹𝗹 𝗶𝘀𝘀𝘂𝗲 here:
ojs.ub.uni-konstanz.de/srm/issue/vi...
#SurveyResearch #SurveyMethods #SurveyClimate #ESRA
Vol. 19 No. 3 (2025): Survey Climate and Trust in Scientific Surveys
| Survey Research Methods
ojs.ub.uni-konstanz.de
October 28, 2025 at 9:09 AM
📘 𝗡𝗲𝘄 𝗦𝗽𝗲𝗰𝗶𝗮𝗹 𝗜𝘀𝘀𝘂𝗲 𝗼𝗳 𝙎𝙪𝙧𝙫𝙚𝙮 𝙍𝙚𝙨𝙚𝙖𝙧𝙘𝙝 𝙈𝙚𝙩𝙝𝙤𝙙𝙨:
“Survey Climate and Trust in Scientific Surveys.”
Explore how the relationship between surveys and society has evolved.
🔗 Read the 𝗳𝘂𝗹𝗹 𝗶𝘀𝘀𝘂𝗲 here:
ojs.ub.uni-konstanz.de/srm/issue/vi...
#SurveyResearch #SurveyMethods #SurveyClimate #ESRA
“Survey Climate and Trust in Scientific Surveys.”
Explore how the relationship between surveys and society has evolved.
🔗 Read the 𝗳𝘂𝗹𝗹 𝗶𝘀𝘀𝘂𝗲 here:
ojs.ub.uni-konstanz.de/srm/issue/vi...
#SurveyResearch #SurveyMethods #SurveyClimate #ESRA
This game is setting football back a bit. 😂🤨
But Jimmy MFin Rolder came to drink milk and kick ass. And the milk is gone.
But Jimmy MFin Rolder came to drink milk and kick ass. And the milk is gone.
October 26, 2025 at 3:01 AM
This game is setting football back a bit. 😂🤨
But Jimmy MFin Rolder came to drink milk and kick ass. And the milk is gone.
But Jimmy MFin Rolder came to drink milk and kick ass. And the milk is gone.
All Michigan had to do was execute and not make this a dogfight. Congrats, you now have a dogfight. 🙄 Get your shxt together.
October 26, 2025 at 12:59 AM
All Michigan had to do was execute and not make this a dogfight. Congrats, you now have a dogfight. 🙄 Get your shxt together.
A reminder that I will be giving this webinar today as part of ICPSR’s #SBECCC. This is the first time that I’ll be introducing this work, which we believe has great potential to improve the quality of imputations in a variety of settings. Would love to hear thoughts! Link below.
Save the date for an upcoming #sbeccc webinar featuring @bradytwest.bsky.social on October 23rd! Learn more: myumi.ch/E8Ap6
October 23, 2025 at 10:39 AM
A reminder that I will be giving this webinar today as part of ICPSR’s #SBECCC. This is the first time that I’ll be introducing this work, which we believe has great potential to improve the quality of imputations in a variety of settings. Would love to hear thoughts! Link below.
Reposted by Brady West
Reposted by Brady West
Micha Fischer, Roderick Little & @bradytwest.bsky.social investigate multiple imputation under missing not at random assumptions w/response indicators as covariates, finding in MNAR data scenarios, methods including RIs can improve performance for both analytic & descriptive inference myumi.ch/n19nk
Multiple imputation under missing not at random: incorporating response indicators into sequential imputation
Multiple imputation (MI) of missing values is mostly applied under the assumption of missing at random (MAR), but the alternative missing not at random (MNAR) assumption may be more plausible. MI a...
myumi.ch
October 20, 2025 at 7:32 PM
Micha Fischer, Roderick Little & @bradytwest.bsky.social investigate multiple imputation under missing not at random assumptions w/response indicators as covariates, finding in MNAR data scenarios, methods including RIs can improve performance for both analytic & descriptive inference myumi.ch/n19nk