Abstract:
|
Too often, researchers think that statistics is something they or statisticians do to data after the data are collected. For decades, statisticians have made in-roads with their scientific colleagues to play a strategic role in design of research studies, which obviously includes considerations beyond sample size calculations. However, there is still more to statistics than design of a study and analysis. This talk will explore "statistical ergonomics" as a more holistic view of statistical thinking that includes the many intermediate steps for handling data, choosing an analysis plan, perspectives on inference beyond the p-value, assessing reliability of results and more. This is especially important in an era where many statistical tools are within easy reach of the casual user, and many more 'data professionals' are entering the field of data analysis without due consideration for statistical thinking.
|