Posted by Kelley MacEwen on 2/9/15 9:00 AM
Read more about me: Biography
To kickoff Summit’s 2015 Professional Education and Knowledge Seminars (PEAKS) series, Dr. Shane Thompson presented on exogeneity and endogeneity, two core concepts of statistical analysis and experimentation.
In order to achieve unbiased results in analysis, statisticians must be mindful of the relationship between independent variables and the outcomes of interest. If the independent variable of interest is chosen or assigned relative to how it will affect the outcome of interest, it is endogenous. Conversely, an independent variable is exogenous when it is assigned randomly (or quasi-randomly).
Endogenous variables lead to biased estimates and can confound otherwise sound analyses. Evaluators, statisticians, and econometricians must be able to identify endogenous and exogenous variables, and implement statistical methods to mitigate endogeneity.
Take a look at his presentation below: