Regression is a statistical technique used to understand multivariate associations between predictor, demographic, prognostic, clinical, confounding variables and outcome variables. The type of regression technique used depends upon the scale of measurement of the outcome variable. Click on a button below to learn more about that type of regression.
Regression analysis for a dichotomous categorical outcome.
Regression analysis for a polychotomous categorical outcome.
Regression analysis for an ordinal outcome.
Regression analysis for survival or time-to-event based on a dichotomous categorical outcome.
Regression analysis for a continuous outcome.
Regression analysis for a count outcome where the mean is higher than the variance.
Regression analysis for a count outcome where the variance is higher than the mean.
Use regression models to generate the probability of a disease state given known risk factors and provide a scoring system for clients and clinicians.
There are three different methods of conducting a regression model. Different methods allow researchers to 1) control for confounding variables (simultaneous regression), 2) choose the best set of predictor variables that account for the most variance in an outcome (stepwise regression), or 3) test theoretical models (hierarchical regression). Click on a button below to learn more.
Regression diagnostics are absolutely necessary when conducting any of the aforementioned regression analyses. Click on a button below to learn more about each type of diagnostic test.
