Practical significance of the difference in means

Abstract

It is shown how the standardised difference (the effect size) between two population means can be used to establish significance when the populations are observed in totality. When dealing with two samples methods are given to determine the practical importance of a statistically significant difference. The usual effect size formula is adapted to deal with cases where populations have different standard deviations. Opsomming Dit word aangetoon hoe die gestandaardiseerde verskil (die effekgrootte) tussen twee populasiegemiddeldes gebruik kan word om beduidenheid t.o.v. volledig waargenome populasies te bepaal. In die geval van twee steekproewe word metodes gegee om die praktiese belangrikheid van 'n statistiese beduidende verskil vas te stel. Die gewone effekgrootte formule word aangepas ten einde gevalle waar populasies verskillende standaardafwykings het te hanteer.