Andrew Gelman, an acclaimed statistician and author, discusses his new book, Active Statistics. He explores the significance of engaging teaching methods that emphasize storytelling and active participation in statistics education. Gelman critiques traditional grading systems in the U.S. and France, highlighting how cultural perspectives shape learning experiences. The conversation also delves into challenges in teaching causal analysis and the importance of innovative strategies in making Bayesian statistics accessible to all.

Join Andrew Gelman, a statistics and political science professor at Columbia, Jennifer Hill from NYU specializing in causal questions, and Aki Vehtari, an expert in computational modeling from Aalto University, as they dive into the enchanting world of regression analysis. They share insights on their writing journey, offer ten tips to enhance regression modeling, tackle the challenges of statistical significance, and reveal the power of storytelling in data education. Plus, there's a whimsical discussion about exploring Mars!

Psychologist tests field of psychology, scientists bet on success-rate, paradoxes of human nature explored. Replication crisis discussed with guest voices, focusing on reproducibility and scientific claims. Challenges of replication studies, questioning research standards and reliability. Impact of multiple comparisons in studies, ethical concerns highlighted. Exploring researcher degrees of freedom and philosophical debates in scientific research.

In this new episode, we talk about the interplay between statistics and data visualization. We do that with Andrew Gelman, Professor of Statistics and Political Science at Columbia University, and Jessica Hullman, Professor of Computer Science at Northwestern University. Andrew started the popular blog “Statistical Modeling, Causal Inference, and Social Science,” which has an active community of readers and has been around for many years. Jessica started contributing lately with many exciting posts, several of which have to do with data visualization. In the episode, we touch upon many topics, including the story behind the blog, the role of surprises, anomalies, and storytelling in science, the Anscombe’s quartet, and exploratory data analysis. Links Jessica Hullman: http://users.eecs.northwestern.edu/~jhullman/ Andrew Gelman: http://www.stat.columbia.edu/~gelman/ Blog: Statistical Modeling, Causal Inference, and Social Science: https://statmodeling.stat.columbia.edu/ Andrew’s 2003 paper on visualization as model checks: “Exploratory Data Analysis for Complex Models” Jessica and Andrew’s follow-up article expanding on the idea of model checks for visualization research: “Designing for Interactive Exploratory Data Analysis Requires Theories of Graphical Inference” Andrew and Thomas’ paper on stories in social sciences: “When Do Stories Work? Evidence and Illustration in the Social Sciences” — Remember: our podcast is listener-supported; please consider donating Using Patreon or Paypal. Thanks! Related episodes Big Data Skepticism w/ Kate CrawfordScience Communication at SciAm w/ Jen ChristiansenStatistical Numbing with Paul SlovicMachine Bias with Jeff LarsonCalling Bullshit with Carl Bergstrom and Jevin WestData Science and Visualization with David Robinson

Experts Andrew Gelman and Zeynep Tufekci discuss the failures of polling in recent elections, questioning its impact on democracy. They delve into the reasons behind polling errors and the need for improvement. Anthony Wells sheds light on the functioning of the polling industry beyond elections.