Journals Information
International Journal of Social Work and Human Services Practice(CEASE PUBLICATION) Vol. 2(3), pp. 101 - 105
DOI: 10.13189/ijrh.2014.020308
Reprint (PDF) (196Kb)
Sherlock Holmes and the Death of the Null Hypothesis
Paul Wilson *
Calibre Global, Australia
ABSTRACT
In the eighty years since R.A. Fisher鈥檚 original work, null hypothesis significance testing has become the ubiquitous research methodology in fields as diverse as biology, agronomy, social science, psychology, epidemiology and most forms of medical research. Throughout this period, the method has been heavily criticised by statisticians who point to a range of problems. The most severe of these is the Bayesian fallacy of wrongly accepting an alternative hypothesis and the problem becomes most obvious in the screening test anomaly whereby a medical screening test can deliver more than 50% of false positives. This paper examines the inverse logical fallacy, that of failing to reject a null hypothesis and thereby accepting an invalid conclusion. A definition of the fundamental concept of experimental completeness leads to a reductio ad absurdum proof. The paper finishes with a live example of the dangerous consequences of accepting an invalid null hypothesis out of a null hypothesis significance test.
KEYWORDS
Null Hypothesis, Evidence, Methodology, Rigour, Rigor, Bayes
Cite This Paper in IEEE or APA Citation Styles
(a). IEEE Format:
[1] Paul Wilson , "Sherlock Holmes and the Death of the Null Hypothesis," International Journal of Social Work and Human Services Practice(CEASE PUBLICATION), Vol. 2, No. 3, pp. 101 - 105, 2014. DOI: 10.13189/ijrh.2014.020308.
(b). APA Format:
Paul Wilson (2014). Sherlock Holmes and the Death of the Null Hypothesis. International Journal of Social Work and Human Services Practice(CEASE PUBLICATION), 2(3), 101 - 105. DOI: 10.13189/ijrh.2014.020308.