Causal Inference & Experiments

Watch

• Tackling a Small Data Search Challenge Using Causal Inference - Liang Wu, ML Data Scientist, Airbnb ↗Lucidworks
• Cambell Causal Model 2 ↗Chuck Smith
• Causal inference and finite population sampling ↗INDIAN INSTITUTE OF MANAGEMENT AHMEDABAD - IIMA
• Causal Inference by Paul R. Rosenbaum · Audiobook preview ↗Google Play Books
• Lessons 7, 8, 9 : Experiments, Observations, and Document Analysis ↗Mark Joseph Calano | Samo't Sari Sa Simula
• London Cholera Outbreak: Was it a Natural Experiment? Causal Inference Bootcamp ↗Mod•U: Powerful Concepts in Social Science

Read

• Causality: Models, Reasoning, and Inference ↗Unravel the mathematical foundations of cause and effect, moving beyond mere correlation to true understanding with this seminal work.Judea Pearl
• Mostly Harmless Econometrics: An Empiricist's Companion ↗Navigate the practicalities of identifying causal links in observational data with this witty and rigorous guide for the empirical researcher.Joshua D. Angrist and Jörn-Steffen Pischke
• Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction ↗Master the potential outcomes framework, the bedrock of modern causal analysis, through a comprehensive and authoritative exposition.Guido W. Imbens and Donald B. Rubin
• Identification of Causal Effects Using Instrumental Variables ↗Grasp how instrumental variables can isolate causal effects, even in the presence of unobserved confounders, a cornerstone of empirical economics.Joshua D. Angrist, Guido W. Imbens, Donald B. Rubin
• Causal Inference in Statistics: An OverviewGain a succinct yet profound understanding of the graphical approach to causality, illuminating the logic behind statistical inference.Judea Pearl

Voices to follow

• Judea Pearl ↗His seminal work on causality, particularly the development of Bayesian networks and the 'do-calculus', provides the mathematical framework for understanding cause and effect in complex systems.Computer Scientist and Philosopher, University of California, Los Angeles
• Guido Imbens ↗A Nobel laureate, he is recognised for his methodological contributions to the analysis of causal relationships, particularly in the context of natural experiments and instrumental variables.Econometrician, Stanford University
• Joshua Angrist ↗Sharing the Nobel Memorial Prize, his empirical and methodological innovations have advanced the use of natural experiments to establish robust causal links in economics and social sciences.Econometrician, Massachusetts Institute of Technology
• Miguel Hernán ↗He is a prominent figure in translating complex causal inference theory into practical, accessible methods for epidemiological and public health research, notably through his 'Causal Inference: What If' textbook.Epidemiologist and Biostatistician, Harvard T.H. Chan School of Public Health

Glossary

• BiasBias is when something unfairly influences the results of an experiment or observation, making them inaccurate or misleading. It can make you think there's a cause-and-effect relationship when there isn't, or hide one that exists. For example, if you only ask your friends about a new video game, and all your friends love that type of game, your survey results might be biased because you didn't ask a variety of people.
• Causal InferenceThis is the process of figuring out if one thing truly causes another, rather than just happening at the same time. It's about being sure that an observed change is because of a specific action or event. For example, if you want to know if a new study method (cause) actually improves grades (effect), you're doing causal inference.
• Cause and EffectA cause is something that makes another thing happen, and the effect is the result of that happening. It's like a chain reaction where one event directly leads to another. For example, if you don't study for a test (cause), then getting a low score (effect) is a direct result.
• Control GroupIn an experiment, the control group is the group that does not receive the special treatment or change you are testing. They serve as a baseline for comparison. For example, if you're testing a new fertilizer, the control group would be plants that get no fertilizer, while other plants get the new one.
• CorrelationCorrelation means two things tend to happen together or change together, but it doesn't necessarily mean one causes the other. They might just be related in some way. For example, ice cream sales and drowning incidents both increase in the summer. They are correlated, but eating ice cream doesn't cause drowning; hot weather causes both.
• Dependent VariableThis is the variable that you measure to see if it changes as a result of the independent variable. It's the "effect" that you are observing. For example, in the plant light experiment, the plant's height or the number of leaves would be the dependent variables because you are measuring how they respond to the light.
• ExperimentAn experiment is a planned test where you change one thing on purpose to see what happens to another thing. It's designed to find out cause-and-effect relationships in a controlled way. For example, a science fair project where you give different amounts of water to plants to see which grows tallest is an experiment.
• Independent VariableThis is the variable that you purposefully change or manipulate in an experiment. It's the "cause" that you are testing to see if it has an effect. For example, if you're testing how different amounts of light affect plant growth, the amount of light is the independent variable because you are controlling it.
• RandomizationRandomization means assigning people or things to different groups completely by chance, like flipping a coin. This helps make sure the groups are similar at the start, so any differences later are likely due to what you changed in the experiment. For example, if you're testing two different types of sports drinks, you'd randomly assign students to drink one type or the other, so you don't accidentally put all the naturally faster runners in one group.
• Treatment GroupThe treatment group is the group in an experiment that receives the specific condition, intervention, or change you are testing. You compare their results to the control group. For example, in a test of a new study app, the treatment group would be the students who use the app, while the control group uses their usual study methods.
• VariableA variable is anything that can change or vary in an experiment or observation. It's a characteristic or factor that can have different values. For example, in an experiment about plant growth, the amount of water, sunlight, or the plant's height are all variables.

Threads 10

Where this connects to other fields, and why it's worth knowing.

• Private Law: Contract, Tort & Property Law

  When a court asks 'did the driver's texting cause the crash?', it uses the 'but-for' test: would the crash still have happened if he hadn't been texting? That means imagining a parallel world where he kept his eyes on the road. Statisticians do the exact same thing to prove cause and effect, comparing our world to the one that didn't happen.

• Design Thinking Arts & Design

  When a designer builds a rough prototype, they're really running a science experiment. The prototype is a guess, 'people will find this easier,' built cheap so you can quickly prove it wrong. That's the same logic as A/B testing, where you try two versions and let real results decide. Good design is testing hypotheses first, and looking pretty second.

• Psychotherapy Psychology

  Compare wildly different kinds of talk therapy and a strange thing shows up: they mostly work about equally well. It's nicknamed the 'dodo bird verdict,' after the dodo who declared 'everybody has won.' That hints the fancy branded technique matters way less than we assume, which is a nightmare for researchers trying to prove their method is the best one.

• Quantum Computing Technology

  Two 'entangled' particles can be split across the galaxy yet always match up perfectly when measured, spookily in sync. But here's the twist: neither one sends any signal to the other, so one can't be causing the other. It's nature's cleanest proof that two things moving perfectly together does not mean one is pulling the other's strings.

• Precision Medicine Health

  A big drug trial tells you a pill helps people on average. But precision medicine points out there's no such thing as the 'average patient' sitting in the room, only this specific person with their own genes. So doctors get caught in a tug-of-war between what worked for the crowd and what will actually work for you.

• Free Will and Responsibility Philosophy

  Saying 'X caused Y' secretly claims that in some other world where X didn't happen, Y wouldn't have either. That imaginary 'other world' is also how we assign blame and praise: he could have chosen differently. So a deep question about free will is quietly baked into every statistics class about cause and effect.

• AI in Medicine Health

  A medical AI that just spots patterns can be confidently wrong the moment doctors change how they treat people, because the patterns it memorized shift too. To be safe it has to learn what actually causes what, not just what usually appears next to what.

• Carbon Markets & Net Zero Environment

  When you pay to protect a forest as a 'carbon offset,' it only counts if those trees really would have been chopped down otherwise. So you're buying a what-if that never happened. And proving a what-if is the hardest thing in all of statistics.

• Environmental and Climate Law Law

  You couldn't sue anyone over a heatwave, because who's to say it wasn't just bad luck? Now "attribution science" can measure how much more likely human pollution made this exact storm. The lawsuit was waiting on the math to catch up.

• Games, Play & Interactive Media Arts & Design

  A randomized evaluation of mathematics video games assigned access to an intervention and measured learning outcomes. This illustrates why comparing people who choose to play with nonplayers is insufficient: random assignment helps separate a game’s effects from prior ability, motivation and other differences between players.

  Sources: Chung, Choi, Baker and Cai — The Effects of Math Video Games on Learning: A Randomized Evaluation Study (2014) ↗