Exploring Logistic Regression

Within statistical analysis, one primary concern is the relationship between the variables involved in a particular study. This can be, at times, the sole purpose of the analysis altogether. An extremely useful tool utilized to show the presence of a relationship between variables and to identify the relation is regression analysis.

The simplest form of regression analysis is linear regression. A more detailed description of linear regression can be found here.

If you need a refresher on linear regression, then it is encouraged that you visit the link above. For our purposes here, we are concerned more with linear regression’s comparable counterpart logistic regression.

In logistic regression the dependant variable is binary or ordinal. Logistic regression is generally used for predicting the probability that a certain event will occur based on numerical or categorical predictor variables. For example, assume a university is conducting research on the effect of Graduate Record Exam (GRE) scores and grade point average (GPA) on admissions into the university’s graduate program. A logistic regression model can then be created from previously observed admissions, which can then provide the basis to predict future admissions.

While today’s software and program can perform logistic regression for you, if you are interested in the math behind it all please explore here.