Maria Gargiulo

blog post: struggles with equivalent weights

you've done MRP.

someone asks you for survey weights.

how to get them ?

blog post: beyond balancing

in midterms, voters tend to support the out party for balance

do polls still help predict midterms ? yes

The LSE and South Coast ESRC Doctoral Training Partnerships, in collaboration with the British Society of Population Studies, would like to invite you to a jointly-organised masterclass entitled: “Being an academic in Population Studies: a masterclass for PhD students on data challenges and career progression” Monday 3 of November 2025, 10am – 4.15 pm (Coffee from 9.30 am, Lunch between 1.15-2.15pm), LSE PhD Academy, LRB 4.02, Lionel Robbins Building (4th Floor), 10 Portugal Street, London WC2A 2HD. If you would like to reserve a space, please sign up here: https://www.eventbrite.co.uk/e/being-anacademic-in-population-studies-a-masterclass-for-phd-students-tickets-1693083350879? aff=oddtdtcreator For DTP-funded students, we expect that funding for travel expenses and/or accommodation will come from their home DTP (ie. RTSG). For non-DTP-funded students, five bursaries will be available. If you wish to apply for a travel bursary for your travel expenses, please send an email to Adrien Allorant with a provisional budget by the 13th of October.

LSE and @bspsuk.bsky.social are hosting a workshop on "Being an academic in population studies" on 3 November! It'll be a nice mix of methods training and career advice with great talks by @ericbschneider.bsky.social, Wendy Sigle, José Manuel Aburto, and others!

www.eventbrite.co.uk/e/being-an-a...

blog post: Fat Bear Week

Basu's Bears is a lesson in:

1) using auxiliary information (pre-salmon-feasting weights)

2) how bad an unbiased estimator can be

statmodeling.stat.columbia.edu/2025/09/23/s...

The data is now available for all teams who've already signed up! More sign-ups very welcome! Happy analysing! 💻📊

blog post: random sampling is not leaving

we turned to response instrument Z because random sampling is "dead"

but does this method still rely on starting with random sampling ?

blog post: random sampling is not leaving

we turned to response instrument Z because random sampling is "dead"

but does this method still rely on starting with random sampling ?

Without Encryption My Work Wouldn’t Be Possible Our work isn’t always popular. It can infuriate those in power who want to cover up incriminating truths about the past and current trends in state violence. So when I think about HRDAG’s threat model...

HRDAG stands firmly with the human rights community in insisting that encryption is not optional — it is fundamental to the defense of truth.

Any attempt to ban or backdoor encryption technology is a threat to the human rights community.

hrdag.substack.com/p/without-en...

¿Qué son los controles civiles y por qué importan para resistir a la militarización? 🪖

Te esperamos en este espacio gratuito y accesible convocado por Intersecta. Participa nuestra analista de datos Sierra Wells.

Regístrate: bit.ly/FM25INTR

📆 Jueves 11 de septiembre, 19 hrs.

blog post on imputation (again):

we want E[Y|X] but X can be missing

@lucystats.bsky.social @sarahlotspeich.bsky.social @glenmartin.bsky.social @maartenvsmeden.bsky.social et al. say:

random imputation should use Y
deterministic imputation shouldn't

statmodeling.stat.columbia.edu/2025/09/09/s...

🚨Job Offer🚨PhD Student Position
The @lshtm.bsky.social, in partnership with the @mpidr.bsky.social is opening an 18-month Research Assistant position, with the successful candidate embarking on a PhD program at LSHTM.
www.demogr.mpg.de/en/career_61...
@ezagheni.bsky.social #phd #jobs #demography

blog post: connections between survey statistics and experimental design.

split-plot designs are analogous to cluster sampling.

blocking is analogous to stratification.

featuring an experiment by Arjun Potter and colleagues at NM-AIST !

Job Opportunity at LSHTM: Research Assistant The London School of Hygiene & Tropical Medicine (LSHTM) is one of the world’s leading public health universities. Our mission is to improve health and health equity in the UK and worldwide; working i...

🗣️Exciting PhD opportunity!

Join @lshtm.bsky.social PSG and @mpidr.bsky.social as a PhD student/Research Assistant on an 3-year project leading to a funded PhD. The role focuses on developing cutting-edge methods to model health & mortality in low-income countries.
jobs.lshtm.ac.uk/Vacancy.aspx...

One Epidemic, Many Estimates (1EME) One Epidemic, Many Estimates (1EME)

📢 Interested in excess mortality methods, and want a challenge? I'm organising the "One Epidemic, Many Estimates" (1EME) project! Register *now* as a many analyst team (submissions due 15 March 2026), and then join us at LSE for a workshop on 21-22 May 2026! (1/n)

www.lse.ac.uk/Economic-His...

blog post: Thomas Lumley writes about Interviewing your Laptop

what are the problems with using LLMs as survey respondents ?

how are these similar to problems with poststratification ?

CC @tslumley.bsky.social

blog post: answers from the BLS

2 weeks ago we learned about the CES employer survey that produces the jobs count.

we asked: why use employment size in stratification but not nonresponse adjustment ?

BLS responded !

statmodeling.stat.columbia.edu/2025/08/19/s...

blog post: 2nd helpings of the 2nd flavor of calibration 🍨🍨

in political surveys, we "logit shift" predictions to match known aggregates (e.g. total Democratic votes).

but what happens for multinomial outcomes ?

a fun excuse to review IPF/raking 🍂

statmodeling.stat.columbia.edu/2025/08/12/s...

blog post: BLS Jobs Report

let's learn about the CES employer survey that produces the jobs count.

late reporting (a form of nonresponse) results in revisions.

my first (naive !) question: why use employment size in stratification but not nonresponse adjustment ?

📢Calling all sociology scholars, including social policy and demography: applications for the 2026-27 Sociology Visitorships at Nuffield College are now open. 🗓️Deadline: 29 Sep 2025

ℹ️More info & applications: www.nuffield.ox.ac.uk/our-research...

blog post: a new paradigm for polling

so far we assumed response R is independent of outcome Y **within X**

but if R can depend on Y, what to do ?

one idea: use a response instrument Z

statmodeling.stat.columbia.edu/2025/07/22/s...

details of the eve

Join us to celebrate five years of #DataAgainstFeminicide
& launch the Swahili version of our tools, co-designed w/ Femicide Count – Kenya

🗓️ Aug 14, 2025
🕚 11AM EDT | 12PM ARG/BRA | 6PM EAT
🌍 In English w/ live interpretation to Spanish, Portuguese, & Swahili
Register: mit.zoom.us/webinar/regi...

blog post about longitudinal/panel data:

panel data includes repeated surveys of the same people over time.

this structure can be incorporated into models using person-level effects.

but misspecifying the person-level effects distribution can cause bias.

blog post about imputation:

With nonresponse worsening, we want to adjust for a lot of covariates.

This often means handling many missing covariates.

In theory, fit one big model for everything. But how can practitioners handle this ?

*really* excited for this !

love the name: Structural Zero.

Alan Agresti's Categorical Data Analysis book offers a good explanation (which I'm sure the amazing authors at @hrdag.org will get into):

blog post about Sparsified MRP:

With nonresponse worsening, we want to adjust for a lot of covariates.

Estimates from such big models will be unstable without a lot of data and/or regularization.

Have you seen MRP with sparsifying priors ?

statmodeling.stat.columbia.edu/2025/07/01/s...

blog post about poststratification:

which estimate is best ?
1. unadjusted sample mean
2. classical poststratification
3. regularized poststratification (e.g. MRP)

statmodeling.stat.columbia.edu/2025/06/24/s...

Finally, this paper is a love letter to my colleagues @hrdag.org and our partners. June 4, 2018, seven years ago today (!), was my first day in the HRDAG office as an intern (!!). So much of my thinking has been shaped by the time between then and now; this paper is a culmination of it.

While the paper focuses primarily on conflict mortality, one of the key strengths of the methods used in the workflow is that they are context agnostic. These methods can easily be adapted to examine other causes of death or other events of demographic interest.

In demonstrating the workflow, I draw heavily on a joint project between the JEP, the CEV, and @hrdag.org about the armed conflict in Colombia that I’ve been collaborating on since 2020.

hrdag.org/jep-cev-colo...

This workflow—consisting of record linkage, imputation, and multiple systems estimation (capture-recapture)—tackles each of these challenges in turn, allowing researchers to go from disparate and statistically biased lists of victims to estimates of the total victim population.