P = nothing, or why we should not teach healthcare students about statistics.
Nurse education today, 31 (8):
837-40 (November 2011)7057<m:linebreak></m:linebreak>CI: Copyright A(c) 2010; JID: 8511379; 2010/08/05 received; 2010/11/08 revised; 2010/11/11 accepted; 2010/12/03 aheadofprint; ppublish;.DOI: 10.1016/j.nedt.2010.11.017
Abstract
Tests of statistical significance are ubiquitous in the nursing and medical literature, and yet they are widely misunderstood and misinterpreted. This paper discusses this, with particular reference to null hypothesis significance testing and the associated p-value. It concludes that the assumptions behind these tests are too stringent for most undergraduate nursing students, and this, alongside the poor understanding of their meaning means that their teaching should have a limited place in the curriculum, which should instead concentrate upon exploratory analysis and understanding measures of effect.