Online Program

Friday, February 19
CS09 Do the Right Thing Fri, Feb 19, 2:00 PM - 3:30 PM
Crystal II

Common Pitfalls and Misconceptions in Statistics with Suggested Solutions (303141)

*Victoria Cox, Dstl 

Keywords: pitfalls, misconceptions, communication, translation, DoE, analysis, graphs, confidence

This presentation will discuss common pitfalls and misconceptions made by both the statistician and the customers when using and interpreting statistical analysis output. Using real-world examples, this presentation will demonstrate that errors can be made in data collection, graphical displays, summary statistics, model choices, communicating results to customers, and the consequences of these errors in the context of decision-making. It will also show common customer misconceptions about sample sizes, the magic 95% confidence level, “getting the right answer from the data,” and the mark one eyeball test ( i.e., statistical inference through physical examination as opposed to statistical analysis). Through case studies, the presentation provides evidence that these pitfalls can be avoided and misconceptions can be overcome.