Statistics in a Nutshell

Chapter 6. Critiquing Statistics Presented by Others

This chapter explains how to read and critique statistics presented by someone else, including those contained in published research articles and workplace presentations, and will teach you to evaluate the research design. You’ll also learn to critique the statistics chosen and their presentation, and common ways authors and presenters try to cover up weaknesses in their data.

The Misuse of Statistics

Broadly, the misuse of statistics falls into two very distinct categories: ignorance and intention. The ignorant use of statistics arises when a person attempts to use descriptive or inferential statistics to support an argument, where the technique, test, or methodology is inappropriate. The intentional misuse of statistics arises when a person attempts to conceal, obfuscate, or over-interpret results that have been obtained. Intuitively, you may think that ignorance arises mostly with complex statistical procedures such as multivariate analysis—and it certainly does—but even basic descriptive statistical procedures are routinely misused.

Intentional misuse is rife in descriptive statistics as well; witness the infamous “stock charting” techniques, in which graph axes are typically labeled with an ordinal scale that manipulates intervals to make a stock appear to rise faster in price than it actually does. The assumptions of inferential testing that make the calculation of statistical tests valid are routinely ignored, as they represent ...