QTM 110 Study Guide - Midterm Guide: Statistical Power, Takers, Random Assignment

• School: Emory University
• Department: Quantitative Theory & Methods
• Course: QTM 110
• Professor: Rodrigo Azuero

Document Summary

Exam 2 study guide by yours truly: threats to experiments. Internal validity threats: chance imbalance, treatment and control group are different by chance even after random assignment, solution: condition x variable/control for x i. Communication/collaboration between treatment and control groups, contaminates treatment. Solution: test for effect on people who are not siblings vs. people who are siblings to determine spillover. Heterogeneity: population has a different (heterogenous) response than the general population. Make sure experimental population is representative, or at least representative of the populations/places you want to apply it. Look at subpopulation to look for heterogeneous effects: Compare treatment to control men/women/poor/educated: regression. Idea of regression: estimate a line that fits the data (short) where. Goal of regression is to control for confounding variables. = + i yi x. = intercept, predicted outcome for individuals with 0 years of education. = how much does y change when x increases by 1.