Follow Regression to the Mean Girls on [⁠Apple Podcasts⁠], [⁠Spotify⁠], or wherever you love to listen!

Christine Zhang⁠ (⁠New York Times⁠) joins ⁠⁠⁠Sarah Lotspeich⁠⁠⁠ and ⁠⁠⁠Lucy D'Agostino McGowan⁠⁠⁠⁠ with a data-driven side quest into the TV show Law and Order. 

Check out the full analysis here: ⁠⁠⁠regressiontothemeangirls.com⁠⁠⁠

Links to content we chat about on the episode:

• ⁠Wrong Number newsletter⁠
• ⁠The Speed of Justice, According to Law & Order⁠

Follow along on instagram: 

⁠@regressiontothemeangirls

Follow Regression to the Mean Girls on [Apple Podcasts], [Spotify], or wherever you love to listen!

⁠Lucy D'Agostino McGowan⁠⁠ and ⁠⁠Sarah Lotspeich⁠⁠ return with a data-driven side quest into the TV show Severance. Lucy analyzes whether Milchick really uses more complicated language than his colleagues.

Check out the full analysis here: ⁠⁠regressiontothemeangirls.com⁠⁠

Links to content we chat about on the episode:

• ⁠⁠Lucy's Severance data R package [mdr]⁠⁠
• ⁠⁠Lucy's other Severance analyses⁠⁠
• ⁠⁠Dunnet⁠⁠
• ⁠⁠Radial plot of who is speaking in Severance ⁠⁠

Follow along on instagram: 

@regressiontothemeangirls

Sarah Lotspeich⁠ and ⁠Lucy D'Agostino McGowan⁠ introduce Regression to the Mean Girls, a podcast about data-driven side quests in pop culture. In this episode, Sarah analyzes 45 recent romance novels to find out how different romance tropes affect the timing of the couple’s first kiss.

Check out the full analysis here: ⁠https://regressiontothemeangirls.com/⁠

Links to content we chat about on the episode:

• ⁠Edward L. Kaplan and the Kaplan-Meier Survival Curve⁠
• ⁠Nonparametric Estimation from Incomplete Observations⁠
• ⁠Not just love, actually: why romance fiction is booming⁠
• ⁠What is the lipstick index?⁠

Follow along on instagram:

@regressiontothemeansgirls

Podcast feed: https://creators.spotify.com/pod/profile/rttmg/

• Emily: ‪@emilyriederer.bsky.social‬
• Ellie: @epiellie.bsky.social
• Lucy: @lucystats.bsky.social

Carly Brantner is an assistant professor of Biostatistics & Bioinformatics at Duke University and Duke Clinical Research Institute.

• multicate: R package for estimating conditional average treatment effects across one or more studies using machine learning methods
• PCORnet® Front Door: Access point for potential investigators, patient groups, and other stakeholders to connect with PCORnet and get support for potential research studies
• Patient-Centered Outcomes Data Repository (PDOCR): De-identified data from 24 (and counting) PCORI-funded studies

• Carly: @carlybrantner.bsky.social
• Ellie: @epiellie.bsky.social
• Lucy: @lucystats.bsky.social

• Vincent's "What is your estimand" section in his {marginaleffects} book: https://marginaleffects.com/chapters/challenge.html#sec-goals_estimand
• Article on defining estimands: https://doi.org/10.1177/00031224211004187
• Andrew's marginal effects post: https://www.andrewheiss.com/blog/2022/05/20/marginalia/
• Andrew's post on "fixed effects" and mariginal effects across different disciplines: https://www.andrewheiss.com/blog/2022/11/29/conditional-marginal-marginaleffects/

• Andrew: @andrew.heiss.phd
• Ellie: @epiellie.bsky.social
• Lucy: @lucystats.bsky.social

In this episode Lucy and Ellie dig into a recently publicized paper, "Vaccination and Neurodevelopmental Disorders: A Study of Nine-Year-Old Children Enrolled in Medicaid", which has gained attention after being promoted by RFK Jr. as evidence that vaccines cause autism.   

Ellie breaks down her Substack critique of the study. Together, she and Lucy discuss the methodological flaws and what a better version of this study might look like.  

• Vaccination and Neurodevelopmental Disorders: A Study of Nine-Year-Old Children Enrolled in Medicaid: https://publichealthpolicyjournal.com/vaccination-and-neurodevelopmental-disorders-a-study-of-nine-year-old-children-enrolled-in-medicaid/
• RFK Jr is promoting a new study claiming "vaccines cause autism" but it doesn't add up. Literally [Ellie's substack]: https://epiellie.substack.com/p/rfk-jr-is-promoting-a-new-study-claiming

Follow along on Bluesky:

• Ellie: @epiellie.bsky.social
• Lucy: @lucystats.bsky.social

🎶 Our intro/outro music is courtesy of Joseph McDade. Edited by Cameron Bopp.

• Vincent's website: https://arelbundock.com/
• Vincent's book "Model to Meaning: How to Interpret Statistical Models With marginaleffects for R and Python": https://marginaleffects.com/

• Vincent: @vincentab.bsky.social
• Ellie: @epiellie.bsky.social
• Lucy: @lucystats.bsky.social

Noah Greifer is a statistical consultant and programmer at Harvard University.

Episode notes:

• WeightIt package: https://ngreifer.github.io/WeightIt/
• MatchIt package: https://kosukeimai.github.io/MatchIt/
• Noah's awesome Stack Exchange post: https://stats.stackexchange.com/a/544958

Follow along on Bluesky:

• Noah: @noahgreifer.bsky.social
• Ellie: @EpiEllie.bsky.social
• Lucy: @LucyStats.bsky.social

🎶 Our intro/outro music is courtesy of Joseph McDade. Edited by Cameron Bopp.

Lucy and Ellie chat about large language models, chat interfaces, and causal inference.

• Do LLMs Act as Repositories of Causal Knowledge?: https://arxiv.org/html/2412.10635v1

Follow along on Twitter:

• The American Journal of Epidemiology: @AmJEpi

• Ellie: @EpiEllie

• Lucy: @LucyStats

🎶 Our intro/outro music is courtesy of Joseph McDade. Edited by Cameron Bopp.