Arbitrarily set at 0.05 (= half of a 10th for statistically ‘significant’ results) when it was first proposed—now frequently extended to 0.001 = a 10th of a 10th of a 10th, for ‘highly significant’ results—this value has turned into a benchmark that divides the numerical continuum into two distinct ranges, overemphasizing the distance between numbers that are above and below this value [ 27 ].
← all excerpts
When bases clash: cultural, cognitive and practical ramifications of numeration systems used in parallel.
1
—
—