ADIA Lab Seminar with Guido Imbens and Geoffrey West, on February 4, 2026

Value of the Information & Strength of the Argument

The talk provides valuable insights into the theory and practice of randomized experiments, grounded in the speaker’s extensive expertise. Imbens effectively argues for the superiority of randomized experiments over observational studies, while acknowledging their limitations and potential for misuse. He supports his points with concrete examples, such as the West Point laptop study and the eBay advertising experiment, which illustrate both the power and the nuances of experimental design. The argumentation is clear and persuasive, though some examples are presented briefly without deep methodological detail.

Scientific Rigor, Source Quality, Title Accuracy

The presentation demonstrates high scientific rigor, consistent with Imbens’s reputation. He references key figures like Fisher, Neyman, and Rubin, and discusses the philosophical debates around experimentation. The sources cited are primarily academic studies and historical references, though specific citations are not always provided in the talk. The title accurately describes the seminar’s content, which is a lecture on randomized experiments. The talk is well-structured and maintains a high level of intellectual integrity.

The title accurately reflects the content, as the seminar features two distinguished speakers, with Imbens focusing on randomized experiments.

8/10

The talk is given by Nobel laureate Guido Imbens, a leading expert in causal inference and randomized experiments. The content is rigorous, well-structured, and based on established scientific principles. However, it is a seminar presentation without peer review, and some claims rely on specific studies that are not fully detailed.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 02:00Introduction to randomized experiments and the fundamental problem of causal inference.
• 10:00Discussion on the importance of defining interventions, with example of child poverty and lottery study.
• 20:00Critique of significance testing and example of data manipulation in a medical trial.
• 30:00Presentation of West Point laptop experiment and its findings on student performance.
• 40:00AI in call centers: experimental evidence on productivity improvement.
• 50:00AI in management consulting: differential effects on routine vs. frontier tasks.
• 1:00:00Remote work experiment at a Chinese travel agency: effects on attrition and satisfaction.
• 1:10:00eBay advertising experiment: demonstrating the value of proper experimentation.
• 1:20:00Introduction to multi-armed bandit designs for efficient treatment selection.
• 1:30:00Dealing with interference: randomizing markets and crossover designs.

• Randomized Experiments, Then and Now — Main topic of the talk by Guido Imbens.
• Hidden Laws of Life, Growth, Death, and Innovation — Title of Geoffrey West's talk, mentioned in the description.

• Rubin Causal Model — Framework for causal inference discussed in the talk.
• Multi-armed bandit — Experimental design approach mentioned for efficient treatment selection.