In the supplementary reading "When Do We Know What We Think We Know," 

http://www.econ.ucla.edu/people/papers/currie/more/ACH.pdf 

1: what is the author Currie's perspective on whether evaluations should be done through random assignment or quasi-experimental methods?  What is she most worried about in study designs? 

2: What types of quasi-experimental methods does she describe, or what types are described in the text? What is the main comparison group for this method? 

3: Do you have a sense that quasi-experimental results have added to knowledge in the field of early childhood?  Do we know more about early childhood education than we did four decades ago? If yes, how so and where do we fall short?