Tuesday

190402

The Cardinal Sins of Skewed Research, Part 2: Racking

Drs. Mary Dan and Michael Eades review the practice of “racking,” one of the cardinal sins of scientific research. In this sin, so termed for its similarity to the practices of medieval inquisitors, scientific data is “tortured until it confesses” a result desired by the researcher (or the researchers’ funder). Among the data manipulations discussed are the selective practices of axis stretching — spacing intervals on a graph to result in a more pronounced visual effect — and the use of relative risk reduction to obscure actual absolute results.

Read MoreThe Cardinal Sins of Skewed Research, Part 2: Racking