Nonexperimental Versus Experimental Estimates of Earnings Impacts

By S. Glazerman, Daniel Levy, David Myers. 2003

370 citations recorded by the source corpus.

Subfields: Economics and Econometrics, Gender Studies, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/4809c13d6c4185aba8ab8dc7cd19c4aab2f52980

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Gender Studies

In Statistics and Probability

Citations in the corpus

Using an Experimental Evaluation of Charter Schools to Test Whether Nonexperimental Comparison Group Methods Can Replicate Experimental Impact Estimates (2012, result): “.......................................................................... 35 B. Sensitivity of Findings to Data Availability, Comparison Group Definitions, and Model Specifications .................................................... 39”
Three conditions under which experiments and observational studies produce comparable causal estimates: New findings from within‐study comparisons (2008, background): “Glazerman, Levy, and Myers (2003) synthesized 12 within-study comparisons meta-analytically while Bloom, Michalopoulos, and Hill (2005) and Smith and Todd (2005) have provided qualitative syntheses.”
Matching With Multiple Control Groups With Adjustment for Group Differences (2008, result): “In particular, previous empirical research (Heckman et al. (1998); Glazerman et al.”
CAN NONEXPERIMENTAL ESTIMATES REPLICATE ESTIMATES BASED ON RANDOM ASSIGNMENT IN EVALUATIONS OF SCHOOL CHOICE? A WITHIN-STUDY COMPARISON (2012, background): “In a meta-analysis of 12 studies that made within-study comparisons, Glazerman, Levy, and Meyers (2003) found that even in analyses using a rich set of covariates and pretreatment outcome measures, nonexperimental methods have “often produced estimates that differed by policy relevant margins” from…”
Standards of Evidence: Criteria for Efficacy, Effectiveness and Dissemination (2005, methodology): “In other words, a few randomized trials will probably provide a more precise answer about whether or not an intervention works than many nonrandomized experiments using equating (e.g., Glazerman et al., 2003).”
Propensity Score Matching (2008, background): “These studies found that PSM could not reliably generate impact estimates in the same direction as well-implemented randomized designs across a range of social programs and policies: reductions in classroom size (Wilde and Hollister 2007); job training and employment services (Glazerman et al. 2003)”
A Survey on Causal Inference (2020, background): “tching procedure, and in this subsection, we briefly discuss what kinds of variables should be included in the matching, a.k.a feature selection, to improve the matching performance. Many literatures [41, 52, 112] suggest to include as many variables that are related to the treatment assignment and ”
An Evaluation of the Chicago Teacher Advancement Program (Chicago TAP) After Four Years. Washington, DC: Mathematica Policy Research (2012, background): “These are all factors that have been associated with more successful (lowerbias) quasi-experimental evaluation designs (Glazerman et al. 2003).”
Another Way Out: The Impact of Juvenile Arrests on High School Dropout (2009, methodology): “I chose one-to-one caliper matching over other PS matching methods because of its relative effectiveness in reducing selection bias (Glazerman et al. 2003), its comparability with the focal matching approach described earlier, and because it is the only PS method that achieved treatment/comparison c”