CRIM 220 Lecture Notes - Lecture 3: Spurious Relationship, External Validity, Internal Validity
Document Summary
Causality and validity in research design sept 16. New charges within 2 years ater date out of prison; similar charges related to original charge; rate of new convicions ater 5 yrs of release; police interacions, etc. Time frame: how you state the problem frames the outcome. Having the right quesion drives everything else. Predicing outcome given some already known characterisics of the individual. Ex. who cheats on tax paying: causal qs. Find out people who have cheated on taxes and gather data and background of them. Ask how a paricular change in one variable would impact another variable. Want to esimate how much iv efect dv. Correlaional v. causal relaionship: correlaional: 2 variables perform in synchronized manner. Variables are correlated: studying increase grade increase. A change in something and a change in another thing as well. One goes up, other goes down: ex. transit pay up = people down, 2. The cause (iv) must occur before the efect (dv)