Quentin Gallea, Ph.D
Think Causally, Act Wisely
A unique perspective
Here is how I can help with your biggest challenges
Detecting and addressing bias (cognitive, behavorial, and statistical) in data/KPIs with Statistics
Improving processes with AI
Tailor-made services to match your needs
"Garbage-in, garbage-out" - a principle especially true in today's data-driven world. This set of trainings ensures that your decisions are based on solid, bias-resistant foundations, by learning practical tools and concepts to maximize the strength of data while addressing the limitations (cognitive, behavourial and statistical bias). Setup the time and date with me and transform the way you interact with data and make decisions that are not mere guesses, but powerful, data-informed steps towards unparalleled success.
Step 1: Robust and Meaningful measures: Data input
Before even going into data analysis, it is key to know what we should measure and focus our effort on. In this section we will follow a checklist to ensure that what we measure is aligned with the goal, will drive the right incentive and reduce the risk of manipulation.
Example: When the state of New York in the 90s started to report publicly on a scorecard the mortality rate of patients from cardiologists, it backfired. This arguably reasonable KPI for surgeons actually led some surgeons to reject patients at risk, the ones requiring help the most, to avoid affecting their score. This particular issue can be easily prevented by knowing the difference between obligation of means and obligation or results and when to apply what, a concept extensively covered in this class.
Step 2: Beyond first impression: Data interpretation and analysis
Once you followed carefully the first step and you started collecting the relevant data, it is time to understand how to interpret adequately this data: what we can say and what we can't. In this section, we are going to see how to use simple and powerful tools to identify and address statistical bias and prevent some cognitive bias.
Example: The mean is the most widely used statistic and a standard when it comes to describing the data. Yet, this simplicity hides useful information and hence makes it often completely misleading. In the worst cases, it is possible that the average represents the inverse of the truth due to a problem called Simpson's Paradox.
Step 3: The causal mindset: Impact assessement
Understanding the concept of causality is crucial from fighting climate change, to our quest for happiness, including strategic decisions making. Unfortunately, it is challenging to assess causality, few people are trained in this niche field and more importantly, this leads to costly mistakes. Hence, I created a set of practical tools allowing you to spot when something is causal or not.
Example: Most of our decisions are based on causal relationships: Does this ads campaign increase our sales? Does working from home increase productivity? Does the change in CEO affect the perception of the clients? Does incentivizing car-sharing reduce emissions related to transportation from our employees?
Data-driven decision-making can be completely misleading if the data itself or the methods are not adequate to answer your questions. More importantly, causality is at the core of decision-making, yet it is incredibly complex to assess.
In the past, I used my expertise in statistics and causal inference (the study of cause and effects) to resolve pressing open questions in a particularly intricate setup: What is the effect of covid lockdowns on the spread of the virus (link to the paper)? What is the effect of sending weapons on the probability and intensity of conflicts in Africa (link to the paper)?
Today, I put my expertise to the service of the industry to help you answer your own pressing questions allowing to minimize the risk of misleading decisions.
I dedicated my career to bridging academia and civil society through innovative teaching and research. Beyond research, I taught statistics and causal inference to more than 10,000 people. over the last decade from undergraduate students in economics, to experienced medical researchers.
After doing scientific research and publishing scientific articles in top peer-reviewed journals (e.g. PNAS, Management Science, Environmental Research Letters, or Journal of Development Economics), I decided to focus my energy on the industry to maximize my impact.
You can download my CV [here]