Carolina Ballroom
Ignorance is … Costly, Not Bliss. So it’s Time to Work Together for Proper Statistical Interpretation (303488)
Nadeesha Lihinikadu, St. Cloud State University*Nancy Kay Sundheim, St. Cloud State University
Most statisticians in the U.S. are aware of the McKinsey report of 2011 that predicts a dire shortage of individuals with deep statistical analysis skills (in the range of 140,000 to 190,000). Need for analytics is high, yet supply is relatively low. Many organizations are doing the best they can but there seem to be many that are accepting people with little to no statistical experience/training to do their statistical analysis. In 2014, statistician Jeff Leek claimed “the vast majority of data analysis is not performed by people properly trained to perform data analysis."
This can result in incorrect conclusions. When describing a probability error known as the Prosecutor’s Fallacy, the New York Times said “… the untrained human mind doesn’t deal with it very well.” As Bill Frost, a Statistical Technical Communications Specialist at Minitab, has pointed out, there are many pitfalls to correct analysis/interpretation. We have selected examples from our own experiences that illustrate this trend of using whoever is available to do the analysis, regardless of their statistical training – or lack of it. But since demand far outweighs supply, organizations often have little choice.