Regression analysis is a statistical framework for modeling relationships between variables to infer or predict outcomes based on inputs within social and economic sciences. It formally distinguishes between metric dependent variables in linear regression models (simple and multiple) versus categorical binary dependent variables in logistic regression, utilizing independent predictors that may be nominal, ordinal, or metric while employing dummy variable encoding for multi-level categories. The theory dictates the selection of model architecture based on the measurement level of the response variable to satisfy specific analytical goals such as quantifying influence or optimizing prediction accuracy.
Prerequisite chain context: requires Scatter Plots in Descriptive Statistics.