Ostalo

Statistical hypothesis testing and some pitfalls

Vesna Ilakovac orcid.org/0000-0002-2560-8800 ; Katedra za biofiziku, medicinsku statistiku i medicinsku informatiku, Medicinski fakultet Sveučilišta J.J. Strossmayer, Osijek, Hrvatska

Sažetak

Data analysis for research purposes usually aims to use the information gained from a sample of individuals in order to make inferences about the relevant population.
Statistical hypothesis testing is a widely used method of statistical inference. It is important to a reader of scientific or expert journals, as well as to a researcher, to understand the basic concepts of the testing procedure, in order to make sound decision and opinion on presented results.
This article gives an overview of basic steps in the general procedure for statistical hypothesis testing and points out some common pitfalls and misconceptions. Questions with particular regard to Pvalue, determining significance level, proving null hypothesis and multiplicity problem were addressed.

Ključne riječi

statistical hypothesis testing; P value; significance level; multiplicity problem

Hrčak ID:

32253

URI

https://hrcak.srce.hr/32253

Datum izdavanja:

5.2.2009.

Podaci na drugim jezicima: hrvatski

Posjeta: 3.553 *