is / are identified by comparing the unique influence of a single independent variable (e.g. one of two ways participants were classified) on an outcome (dependent) variable while temporarily ignoring other independent variables.