POLS 3650 Lecture Notes - Lecture 21: Multivariate Statistics, Spurious Relationship, Confounding
21 Mar 2017
School
Department
Course
Professor
Document Summary
Bivariate analysis misrepresents reality when it omits an important third variable (omitted variable bias) This lecture will discuss what happens when we introduce a control variable (cv) to see if the original relationship we established is accurate. Adding more variables gives us a richer understanding and makes sure nothing is omitted. This brings us into the realm of multivariate statistics. The most straightforward way to control for a third variable is what could be called the. Investigate relationship for cases with value 1 on the control variable. Investigate relationship for cases with value 2 on the control variable. Replication: a scenario where the iv and control variable are completely unrelated, so, the relationship between the iv and dv is replicated after controlling for the. Iv and cv are related to the dv third variable: this is the ideal scenario where there is no omitted variable bias because the third variable has no influence.
