regressiondiscontinuitydesign-陕西师范大学教育经济研究所.pdf

Part 2: Quasi-Experimental Methods (What to do when we can’t do an experiment) 准实验方法 （不能做实验时，我们怎么做？） Sean Sylvia C3e Econometrics Workshop 2015 Outline 课程大纲 1. Quick refresher on what is meant by Quasi- experimental methods 快速复习：什么是准实验方法？ 2. Preliminaries for second part of course 第二部分课程的基础知识 3. Quasi-experimental Topic 1: Regression Discontinuity 准实验专题1：断点回归 PART 2: Impact Evaluation using Quasi- experimental Methods 第二部分：用准试验方法得到的影响评估 W ╨(Y (0), Y (1)) | X i i i i or W ╨(Y (0), Y (1)) | X i i i i RCTs are not always appropriate/feasible: 当RCT不适合/不可实现时： • Effect of a road on access to off farm employment 道路对非农就业的影响 • An education program that already has been rolled out 一个已经实施的教育项目 • Impact of the decisions of migrant families to: leave kids behind (get educated in villages’ rural public schools while living with Grandma … or go with Mom and Dad and get educated in the city in a private, unregulated migrant school). 农民工家庭决策的影响：留守儿童（和奶奶住在 一起在农村公共学校读书或是）...或是和妈妈爸 爸住在一起在城市里私人的不受监管的打工子弟 学校读书 Methods we’ll cover: 我们会用到的方法： • Given W ╨(Y (0), Y (1)) | X assumption (conditional i i i i independence) 假定W 和Y (0),Y (1)在Xi 的条件下相互独立 i i i 1. Matching methods匹配法 • W ╨(Y (0), Y (1)) | X assumption i i i i 假定W 和Y (0),Y (1)是在Xi 的条件下，相互不独立 i i i 1. Difference-in-Differen