About

Who I am

My name is Antoine Soetewey. I hold a PhD in statistics from UCLouvain. I then worked as a postdoctoral researcher in data science and statistics at HEC Liège and UCLouvain Saint-Louis Brussels, before turning to datanalyze full time.

My thesis dealt with the waiting period in mortgage insurance for cancer survivors — a subject at the intersection of biostatistics, multi-state modelling and a very concrete social issue: the right to be forgotten.

I also teach statistics and probability as an invited lecturer at UCLouvain and UNamur.

Under the name datanalyze, I make that expertise available to researchers, professionals and companies who have data to analyse and nobody to do it.

Background

• PhD in statistics — UCLouvain, 2024
• MSc in econometrics — Maastricht University, 2016
• MSc in economics — KU Leuven, 2015
• Exchange program — University of Illinois at Urbana-Champaign, 2014
• BSc in economics — UCLouvain, 2013

That path explains how I work: I came to statistics through applied economics. So I am used to starting from a real-world question rather than a model, and to judging a method by what it lets you conclude rather than by its sophistication.

Research and teaching

My research covered biostatistics, data science and the development of methods for analysing complex datasets. I have published in journals including European Actuarial Journal, Biostatistics & Epidemiology, Vaccine, Open Forum Infectious Diseases, Nutrients and SoftwareX, and taken part in applied research projects in economic microsimulation and automated information processing.

On the teaching side, I currently give courses in probability and statistical inference at UNamur and biostatistics at UCLouvain, after supporting some fifteen different courses as a teaching assistant: descriptive and inferential statistics, data science with R, survival analysis, econometrics, generalised linear models, etc.

That teaching is not a side activity: it is what guarantees I can explain a method to someone who is not a statistician. Which is precisely what a statistical consultant is for.

What I believe

A simple method you understand beats a sophisticated one you merely endure. If you cannot explain why you used a given test, you will not be able to defend it — and an indefensible result is worth nothing, even when it is correct.

Statistics should be accessible. That is what led me to run a blog on statistics and R for years, read by tens of thousands of people each month, and to give private lessons to students, alongside university teaching.

Work is judged by what remains of it. You should leave with the scripts, the documentation and the understanding needed to carry on without me. A successful assignment is one where you no longer need me.

Saying no is part of the job. If your data cannot answer your question, or if your need belongs to a different profession than mine, I will say so. That is more useful to you and more sustainable for me.

Working together

I work in French or English. The analysis is done remotely, so there is no geographic constraint, and I work for clients in Belgium and abroad alike.

You will find what each engagement covers on the services page and my training catalogue.

For the research and teaching side, my academic site is antoinesoetewey.com.