W11: Statistical literacy: evaluating evidence from observational studies


Milo Schield (Augsburg University)


Schedule

Thursday May 16, 1:00 pm – 4:30 pm

Abstract

This workshop will cover the following ten topics:

  • A1. Three kinds of students taking statistics.   
  • A2. Four kinds of influence on a statistic. 
  • A3. Five kinds of studies. 
  • A4. Three kinds of grammar for distinguishing association from causation.   
  • B5. Two biggest statistical problems in dealing with big data.  
  • B6. Five ways to introduce multivariate thinking and confounding.  
  • B7. Two ways to avoid creating cynics when introducing confounding: Fisher vs. Cornfield.  
  • B8. Evaluating the strength of evidence given an association based on observational data.   
  • C9. (20 minutes) Separating statistical literacy from Decision Making. (What not to say about causation given evidence from observational studies)  
  • C10. (30 minutes) Reviewing the first part of Judea Pearl's book: "The Book of Why": the most important book for statistical educators published in decades.

Encore: Evaluating p-values in clinical trials.  

Approach: Three 48 minute sessions. 12 minutes per topic (4 topics per hour)
Extensive audience participation and voting on results for most of the topics.