Beyond chi-square: Is it a fluke?

  • Event: 2016 CAR Conference
  • Speakers: Jennifer LaFleur of Investigative Reporting Workshop; Olga Pierce of University of Nebraska - Lincoln; John Templon of BuzzFeed News; Sarah Cohen of ASU Walter Cronkite School of Journalism; Mark Hansen of Columbia Journalism School
  • Date/Time: Saturday, Mar. 12 at 9:00am
  • Location: Colorado G-J
  • Audio file: Only members can listen to conference audio

**Moderated by Sarah Cohen, The New York Times

Sometimes it's important to know whether a pattern could be a fluke. Is a stockbroker lucky or trading on insider knowledge? Did a player throw a game or have a bad streak? Is the lottery rigged or could the same person win twice?

Reporters often turn to traditional methods like a chi-square test to determine statistical significance, but other techniques, from poisson distributions to monte carlo simulations, might be a better fit. This panel will walk you through some of the methods you can use to answer the question: is it a fluke?

Speaker Bios

  • Sarah Cohen is the Knight Chair in Journalism at the Walter Cronkite School at ASU. Previously, she worked as the editor of a data reporting team at The New York Times focused on long-term enterprise and investigative stories, and as a database editor for The Washington Post. Her awards include the Pulitzer Prize in Investigative Reporting, the Goldsmith Prize and the IRE medal. She is a past president of IRE, and served on the board for eight years.

  • Mark Hansen is a Professor in the Columbia Journalism School where he also serves as director of the Brown Institute for Media Innovation. Prior to joining Columbia, he was a professor in the Department of Statistics at UCLA. Hansen began his career as a Member of the Technical Staff at Bell Laboratories in Murray Hill, NJ. He holds a BS in Applied Math from UC Davis, and a PhD and MA in Statistics from UC Berkeley. @cocteau

  • Jennifer LaFleur is the Investigative Reporting Workshop's data editor and teaches at American University. She previously was a senior editor at Reveal from The Center for Investigative Reporting, where she managed a team of data journalists, investigative reporters and fellows. She is the former data journalism director at ProPublica and has held similar roles at newspapers. She is a former IRE training director and serves on IRE's board.

  • Olga Pierce is on the faculty at the College of Journalism and Mass Communications at the University of Nebraska-Lincoln, and a Data Journalism Adviser to ProPublica. Previously, she worked at ProPublica for ten years as a reporter and Deputy Data Editor. She is a graduate of the Stabile Investigative Seminar at Columbia University. @olgapierce

  • John is an investigative data journalist for BuzzFeed News. He did the data analysis for BuzzFeed's award-winning series on match-fixing in tennis. Prior to joining BuzzFeed John worked at Chicago-based technology startup Narrative Science. @jtemplon

Related Tipsheets

  • Beyond Chi Square
    Sometimes it's important to know whether a pattern could be a fluke. Is a stockbroker lucky or trading on insider knowledge? Did a player throw a game or have a bad streak? Is the lottery rigged or could the same person win twice? Reporters often turn to traditional methods like a chi-square test to determine statistical significance, but other techniques, from poisson distributions to monte carlo simulations, might be a better fit. This PowerPoint presentation will walk you through some of the methods you can use to answer the question: is it a fluke?