Binary data


The forms that Binary Outcomes can take are numerous, but a single statistical characterization is possible.


In linear regression, perhaps the most important concept is conditional means.


Although it would be nice if we could use linear regression with binary outcomes, there are reasons the problem of modeling binary outcomes does not fit in the framework of the linear model.


Nonetheless, conditional means will play a central role in modeling binary outcomes.


The slides used in the videos are found here