Reference no: EM131342

Understand, interpret, and be critical of the concepts of statistical signi cance and p-values.

Statistical signi cance is about whether our sample contains enough information to detect that an eect exists (e.g. to detect that the Binomial parameter p = p0). It says nothing about the size of the eect (i.e. how big is the dierence between p and p0).

Statistical signi cance depends heavily on the sample size, n. An eect of a given size (e.g. the dierence between p0 = 0:5 and p = 0:6) will be detectable in a large sample, but not detectable in a small sample.

 Statistical signi cance is not the same as real-world signi cance. Statistical signi cance does NOT tell us about how large or important a result is. It just calculates OUR evidence in OUR sample for whether or not an eect exists.