01:920:311 Lecture Notes - Lecture 18: Random Assignment, Dependent And Independent Variables, Experiment
29 Nov 2018
School
Department
Course
Professor
Document Summary
Critique folder: will be on the exam, there is a folder on sakai with examples. Do you think these relationships are causal or not: dependent: likelihood of being called back for job interview, academic performance depends on gender empirical association. If you don"t meet all, but only meet certain criteria. It has all the criteria expect random assignment: not ideal, but there is some scientific value and certain justifications for doing them. Introduce other comparison cases: can"t randomly determine what sates someone is in, can compare connecticut to other states, figure 3. Control states selected on basis of having similar demographic and weather conditions: connecticut level of fatalities lower expect for the year before intervention. Its not possible to have a representative sample of us population and try to get them to go to your lab: actual sample is convenient sample of students near by, often highly artificial, making external validity questionable.
