Dichotomizing continuous predictors in multiple regression: a bad idea.
In medical research, continuous variables are often converted into categorical variables by grouping values into two or more categories. We consider in detail issues pertaining to creating just two groups, a common approach in clinical research. We argue that the simplicity achieved is gained at a cost; dichotomization may create rather than avoid problems, notably a considerable loss of power and residual confounding. In addition, the use of a data-derived 'optimal' cutpoint leads to serious bias. We illustrate the impact of dichotomization of continuous predictor variables using as a detailed case study a randomized trial in primary biliary cirrhosis. Dichotomization of continuous data is unnecessary for statistical analysis and in particular should not be applied to explanatory variables in regression models.
|Title:||Dichotomizing continuous predictors in multiple regression: a bad idea.|
|Keywords:||Age Factors, Albumins, Antimetabolites, Azathioprine, Bilirubin, Cholestasis, Data Interpretation, Statistical, Humans, Liver Cirrhosis, Biliary, Randomized Controlled Trials as Topic, Regression Analysis|
|UCL classification:||UCL > School of Life and Medical Sciences
UCL > School of Life and Medical Sciences > Faculty of Population Health Sciences
UCL > School of Life and Medical Sciences > Faculty of Population Health Sciences > MRC Clinical Trials Unit at UCL
Archive Staff Only