Estimating the reproducibility of psychological science

Brian A. Nosek, Science, Aug 31, 2015
Commentary by Stephen Downes

I would not be surprised to see similar results in educational research as well. As the abstract of this study states, "Ninety-seven percent of original studies had statistically significant results. Thirty-six percent of replications had statistically significant results." Of course the reason for this is a mixture of small sample sizes, disproportional representation, and statistical variability. It shows again that we can't rely on (nor cite) individual studies as proof of anything. Especially when they are 'a class of students at a midwestern university'. Academica has the full rundown of links to this widely covered story: CBC | National Post | New York Times | Inside Higher Ed.

Views: 0 today, 217 total (since January 1, 2017).[Direct Link]