6/22/2023 0 Comments How can you lie with statisticsSometimes a researcher might (again, unwittingly) present a statistically significant finding as a big deal, even if it is, in fact, simply spurious and the result of the researcher conducting too many analytical tests. If you conduct 1,000 significance tests, you are likely to find a few of them as significant simply by chance alone. That simply means a result erroneously comes out as statistically significant. In short, the more significance tests someone conducts during the analytical stage, the more likely he or she is to find a spurious result. It often results from a scenario in which a researcher conducted a plethora of significance tests as part of the analytical process. This kind of outcome is actually more common than one might think. The book is written in a very simple manner for. The second source of its success is the authors style of writing. Provocative title at once made this book very recognizable and discussed. Without getting too bogged down in the details, Type-I Error exists when a researcher finds that one result is statistically significant but, in fact, the reality is that the findings do NOT generalize to the broader population of interest. The author says directly that statistics lie, though scientists have been trying to assure people that they can believe the data obtained with its help.
0 Comments
Leave a Reply. |