Parallel session

Skills and tools

3 July

10:10 to 11:20

STCC, Room 3BC




Seeking elusive truths: How to judge statistical results as a non-statistician

Many published papers use limited data sets for making relatively strong scientific claims. Because such claims are usually based on statistical evidence, assessing research requires a basic knowledge of statistics. In this hands-on workshop, we will show that for making a first judgment, in-depth knowledge of the details of statistical analysis may not be necessary. Common sense may be sufficient. The introductory talks will explain basic statistical concepts and commonly used statistical methods. We will then hand out examples from the published literature and discuss them with the attendees. We will emphasize that statistical methods are not barometers of truth, but rather tools for describing the results at hand. Further, journalists should be aware that misunderstandings about statistics are rampant among scientists. This workshop will help journalists understand the strengths and limits of common statistical methods and why garnering "truth" from statistical methods is rarely if ever warranted.