Premium
A researcher's guide to regression, discretization, and median splits of continuous variables
Author(s) -
Rucker Derek D.,
McShane Blakeley B.,
Preacher Kristopher J.
Publication year - 2015
Publication title -
journal of consumer psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 4.433
H-Index - 110
eISSN - 1532-7663
pISSN - 1057-7408
DOI - 10.1016/j.jcps.2015.04.004
Subject(s) - continuous variable , discretization , variable (mathematics) , regression , linear regression , regression analysis , normative , variables , econometrics , variance (accounting) , statistics , mathematics , computer science , economics , epistemology , mathematical analysis , philosophy , accounting
We comment on Iacobucci, Posavac, Kardes, Schneider, and Popovich (2015) by evaluating the practice of discretizing continuous variables. We show that dichotomizing a continuous variable via the median split procedure or otherwise and analyzing the resulting data via ANOVA involves a large number of costs that can be avoided by preserving the continuous nature of the variable and analyzing the data via linear regression. As a consequence, we recommend that regression remain the normative procedure both when the statistical assumptions explored by Iacobucci et al. hold and more generally in research involving continuous variables. We also discuss the advantages of preserving the continuous nature of the variable for graphical presentation and provide practical suggestions for such presentations.