Logistic regression

  • Logistic regression is an example of statistical modelling, used when studying an outcome that has two categories, for example, present/absent, yes/no.

    It is used when this kind of two-category outcome may depend on a number of factors, usually themselves categorical. For example, logistic regression can be used to study how an individual’s sex, age group, social class and other measures might influence their risk of having a particular medical condition. Using logistic regression model allows the effects of all these factors to be studied simultaneously rather than one at a time.

    The results of a logistic regression are usually expressed in terms of odds ratios.