Elizabeth Stuart

Elizabeth Stuart

@lizstuart.bsky.social

Statistician; Professor and Chair @JHUBiostat @JohnsHopkinsSPH, w/links to @SREESociety, @AmericanHealth. Oh, & spouse, mom, runner, traveler.

5.8K followers329 following67 posts

Elizabeth Stuart
@lizstuart.bsky.social

Are you a journalist working on public health topics and wondering how to assess research studies and distinguish correlation from causation? Applications for our free one-day training in DC (2/26/27) now open, with travel funds available! publichealth.jhu.edu/departments/...

publichealth.jhu.eduMastering Causation vs. Correlation: Training for Journalists | Johns Hopkins Bloomberg School of Public HealthA one-day free training at the Johns Hopkins Bloomberg Center in Washington, D.C., open to professional reporters who cover public health.
Elizabeth Stuart
@lizstuart.bsky.social

I’m so glad to see more attention being paid to the data being used to train AI models. The results will only be as good as the data going in.

Alan Regenberg@aregenberg.bsky.social

Dozens of AI disease-prediction models were trained on dubious data. The models are designed to predict someone’s risk of diabetes or stroke. A few might already have been used on patients. www.nature.com/articles/d41...

nature.comDozens of AI disease-prediction models were trained on dubious dataThe models are designed to predict someone’s risk of diabetes or stroke. A few might already have been used on patients.
Elizabeth Stuart
@lizstuart.bsky.social

Looking forward to this; the next in our series on methods for mental health services research! Join us for free!

Johns Hopkins ALACRITY Center @jhalacritycenter.bsky.social

Join us 3/26/26 from 10:30a to Noon to hear Dr. John Jackson present, "Using Causal Decomposition Analysis to Identify Points of Intervention to Reduce Mental Health Disparities" at the next Advanced Methods for Mental Health Services Research Webinar. Register Here: jhjhm.zoom.us/webinar/regi....

Elizabeth Stuart
@lizstuart.bsky.social

Loved this episode of reflection, especially @publichealthpod.bsky.social as a podcast asking and answering questions, with experts from a range of fields who are consider multiple angles and dig into the evidence. Conversations and sources of information like this are more important than ever.

Public Health On Call Podcast@publichealthpod.bsky.social

Hosts @sdesmon.bsky.social, @drjoshs.bsky.social, and @lindsaysmithrogers.bsky.social reflect on 1,000 episodes of the show, the challenges of covering nuanced health topics, and what issues they want to focus on next. 🎧 podcast.publichealth.jhu.edu/1000-celebra...

Elizabeth Stuart
@lizstuart.bsky.social

A great opportunity for folks interested in data science + AI; lots going on at Hopkins in this area! Spread the word and apply!

Johns Hopkins Data Science and AI Institute@hopkinsdsai.bsky.social

Join us in advancing data science and AI research! The Johns Hopkins Data Science and AI Institute Postdoctoral Fellowship Program is now accepting applications for the 2026–2027 academic year. Apply now! Deadline: Jan 23, 2026. Details and apply: apply.interfolio.com/179059

Elizabeth Stuart
@lizstuart.bsky.social

Sharing for awareness of my education research crowd friends -- spread the word and apply!

Johns Hopkins University School of Education@jhueducation.bsky.social

A 16-member committee, chaired by Provost Ray Jayawardhana, will lead a global search for the next dean of the Johns Hopkins School of Education. Dean Chris Morphew to conclude his nine-year term on June 30. https://hub.jhu.edu/2025/12/04/school-of-education-dean-search-committee/

hub.jhu.eduSearch committee formed to identify next Johns Hopkins School of Education dean16-member group chaired by Provost Ray Jayawardhana will lead global search for leader to succeed Chris Morphew
Elizabeth Stuart
@lizstuart.bsky.social

My attempts to help summarize why it's so hard to study things like whether there's a causal link between acetaminophen + autism.

Public Health On Call Podcast@publichealthpod.bsky.social

In a new piece for @statnews.com, @lizstuart.bsky.social explores the challenges of studying cause and effect. She recently joined us for a related conversation on causal relationships between Tylenol and autism 👇 🎧 podcast.publichealth.jhu.edu/953-interpre... ✏️ www.statnews.com/2025/10/23/c...

Elizabeth Stuart
@lizstuart.bsky.social

Starting in a couple of hours; join us!

Johns Hopkins ALACRITY Center @jhalacritycenter.bsky.social

Attend the Johns Hopkins ALACRITY Center’s Advanced Methods for Mental Health Services Research Webinar Series on September 17th from 10:30 a.m. - Noon. Michael Rosenblum, PhD, MS and Joshua Betz, MS will discuss covariate adjustment in randomized trials. jhjhm.zoom.us/webinar/regi...

Elizabeth Stuart
@lizstuart.bsky.social

So fun to come across this! It has been an honor and joy to work with Grace on topics that include proximal causal inference, electronic health records, and measurement error. She is a careful and excellent researcher, and an amazing team member!

Johns Hopkins Dept of Biostatistics@jhubiostat.bsky.social

Grace Ringlein is a PhD student in @jhubiostat.bsky.social focusing on causal inference, measurement error, EHR data, mental health, and algorithmic fairness. Learn more about Grace's work with researchers across multiple @johnshopkinssph.bsky.social departments in our Student Spotlight:

publichealth.jhu.eduStudent Spotlight: Grace Ringlein | Johns Hopkins Bloomberg School of Public HealthGrace Ringlein is a fourth-year PhD student in the Department of Biostatistics whose areas of focus include causal inference, measurement error, Electronic Health Record (EHR) data, mental health, and...
Elizabeth Stuart
@lizstuart.bsky.social

I am grateful for this coverage. Fiona is a friend and sadly just one example of many devoted public servants (who I know personally to be deeply committed, smart, and caring) whose expertise we are currently losing from the government.

Joshua Weitz@joshuasweitz.bsky.social

Profile and interview of Fiona Havers by @apoorvanyt.bsky.social now out in the NYT. A must-read on the loss of experts and expertise. 🎁 link www.nytimes.com/2025/06/18/h...

nytimes.comWhy a Vaccine Expert Left the C.D.C.: ‘Americans Are Going to Die’
Elizabeth Stuart
@lizstuart.bsky.social

@noahgreifer.bsky.social was my post-doc and I cannot recommend him more highly if you need a super smart statistical consultant / programmer. He is the force behind MatchIt, cobalt, and other packages, and is also just a fantastic team member.

Noah Greifer@noahgreifer.bsky.social

Starting to look like I might not be able to work at Harvard anymore due to recent funding cuts. If you know of any open statistical consulting positions that support remote work or are NYC-based, please reach out! 😅

Elizabeth Stuart
@lizstuart.bsky.social

Excited for what will be a flash trip to New England in mid-June for the AI + precision medicine conference in Portland, ME, followed by the Society for Epi Research in Boston! Join me at both!

Elizabeth Stuart
@lizstuart.bsky.social

Great way to learn the basics of policy trial emulation and the importance of careful study design for health policy!

Johns Hopkins ALACRITY Center @jhalacritycenter.bsky.social

If you missed the latest Advanced Methods for Mental Health Services Research Webinar, "Target Trial Emulation for Evaluating Mental Health Policy" from Nicholas J. Seewald, Ph.D., don't worry! You can watch it on the JH ALACRITY Center's YouTube channel: youtu.be/DAXfp8X9ba8?...

youtu.beTarget Trial Emulation for Evaluating Mental Health PolicyYouTube video by JHU ALACRITY Center
Elizabeth Stuart
@lizstuart.bsky.social

This is so important. And especially in non experimental studies where bias - not variance - is the first order concern.

Adam Kucharski@adamjkucharski.bsky.social

The larger the dataset, the larger the false sense of confidence - if bias is baked in, size just makes a flawed measurement more convincing. Xiao-Li Meng has called it the big data paradox: 'The bigger the data, the surer we fool ourselves.' In other words, scale isn’t a substitute for scrutiny.

projecteuclid.orgStatistical paradises and paradoxes in big data (I): Law of large populations, big data paradox, and the 2016 US presidential electionStatisticians are increasingly posed with thought-provoking and even paradoxical questions, challenging our qualifications for entering the statistical paradises created by Big Data. By developing measures for data quality, this article suggests a framework to address such a question: “Which one should I trust more: a 1% survey with 60% response rate or a self-reported administrative dataset covering 80% of the population?” A 5-element Euler-formula-like identity shows that for any dataset of size $n$, probabilistic or not, the difference between the sample average $\overline{X}_{n}$ and the population average $\overline{X}_{N}$ is the product of three terms: (1) a data quality measure, $\rho_{{R,X}}$, the correlation between $X_{j}$ and the response/recording indicator $R_{j}$; (2) a data quantity measure, $\sqrt{(N-n)/n}$, where $N$ is the population size; and (3) a problem difficulty measure, $\sigma_{X}$, the standard deviation of $X$. This decomposition provides multiple insights: (I) Probabilistic sampling ensures high data quality by controlling $\rho_{{R,X}}$ at the level of $N^{-1/2}$; (II) When we lose this control, the impact of $N$ is no longer canceled by $\rho_{{R,X}}$, leading to a Law of Large Populations (LLP), that is, our estimation error, relative to the benchmarking rate $1/\sqrt{n}$, increases with $\sqrt{N}$; and (III) the “bigness” of such Big Data (for population inferences) should be measured by the relative size $f=n/N$, not the absolute size $n$; (IV) When combining data sources for population inferences, those relatively tiny but higher quality ones should be given far more weights than suggested by their sizes. Estimates obtained from the Cooperative Congressional Election Study (CCES) of the 2016 US presidential election suggest a $\rho_{{R,X}}\approx-0.005$ for self-reporting to vote for Donald Trump. Because of LLP, this seemingly minuscule data defect correlation implies that the simple sample proportion of the self-reported voting preference for Trump from $1\%$ of the US eligible voters, that is, $n\approx2\mbox{,}300\mbox{,}000$, has the same mean squared error as the corresponding sample proportion from a genuine simple random sample of size $n\approx400$, a $99.98\%$ reduction of sample size (and hence our confidence). The CCES data demonstrate LLP vividly: on average, the larger the state’s voter populations, the further away the actual Trump vote shares from the usual $95\%$ confidence intervals based on the sample proportions. This should remind us that, without taking data quality into account, population inferences with Big Data are subject to a Big Data Paradox: the more the data, the surer we fool ourselves.
Elizabeth Stuart
@lizstuart.bsky.social

Great way to learn the basics of policy trial emulation for free! Paper here: pubmed.ncbi.nlm.nih.gov/39374529/

pubmed.ncbi.nlm.nih.govTarget Trial Emulation for Evaluating Health Policy - PubMedTarget trial emulation is an approach to designing rigorous nonexperimental studies by "emulating" key features of a clinical trial. Most commonly used outside of policy contexts, this approach is als...
Johns Hopkins ALACRITY Center @jhalacritycenter.bsky.social

Kicking off the Johns Hopkins ALACRITY Center’s Advanced Methods for Mental Health Services Research Webinar Series! Nicholas J. Seewald, Ph.D. will present, "Target Trial Emulation for Evaluating Mental Health Policy" on 3/19/2025 (11a.m-12:15p.m.). Register: jhjhm.zoom.us/webinar/regi...)

jhjhm.zoom.usWelcome! You are invited to join a webinar: ALACRITY Center’s Advanced Methods for Mental Health Services Research Webinar Series. After registering, you will receive a confirmation email about joinin...Topic: Target Trial Emulation for Evaluating Mental Health Policy; Speaker: Nicholas J. Seewald, PhD
Elizabeth Stuart
@lizstuart.bsky.social

This has been one of the most meaningful and rewarding (and sobering) collaborations I have been involved in, especially the integration of advanced stat methods and deep substantive expertise. Crucial empirical data and results that we hope can inform policy discussions.

JAMA@jama.com

🧵 US states that implemented abortion bans saw higher than expected infant mortality rates, with larger increases among Black infants and those in southern states, according to this analysis of US national vital statistics data from 2012–2023. ja.ma/4aVchPn #MedSky

Elizabeth Stuart
@lizstuart.bsky.social

Thankful to be heading into a break but rejuvenated by 3 days talking generalizability @ICERM @BrownUniversity, inc in-person time with post-docs who work remotely and with many colleagues who have become friends.

You are up to date.