Abstract:
|
Confounding is playing a pivotal role in the evolution of statistical literacy. This paper reviews the status of confounding in some of the leading articles on statistical literacy in three time periods: the confounder absent years (1997-2005), the confounder-present years (2013—2016) and the subsequent confounder-absent years (2017-now). This paper provides explanations for the prior omission of confounding, the dramatic increase and the subsequent lapse. This paper argues that we need agreement on the necessary components for a statistical literacy course and we need a way to distinguish a confounder-based statistical literacy course. This paper proposes that the phrase "confounder-based" statistical literacy course be reserved for a statistical literacy course in which confounding is featured at least as often as statistical inference.
|