Home Sections Video ParticipantsProgram Anthony Bryk Video Cecilia Rouse Video A. Thomas McLellan Video Panel Video
SREE - Advancing Education Research
Navigation Bar
Navigation Bar
spacer
spacer Conference Concluded - View Next conference information

Conference Videos
spacer
spacer side image
spacer
spacer spacer spacer
workshops

Want to add a workshop to a completed conference registration? Contact SREE.
All workshops run from 9:00 AM - 12:00 PM on Thursday, March 4th, 2010.

Using Modern Regression Discontinuity Analysis To Measure Effects of Educational Interventions
Howard S. Bloom, MDRC
This workshop (based on a recent book chapter by Dr. Bloom) will introduce participants to the theory and practice of modern regression discontinuity analysis, which is arguably the strongest quasi-experimental design that exists for measuring intervention effects. Developed by Donald Campbell in the early 1960s and applied to numerous education studies during that period, regression discontinuity analysis was largely forgotten until it was rediscovered by economists in the 1990s. Since then, the method has been greatly expanded and is gaining popularity rapidly for intervention research. Although the basic idea of regression discontinuity analysis is extremely simple (which is part of its appeal), modern methods for its application have become rather complex. It is therefore important for potential users to: (1) understand the method’s alternative estimation procedures, (2) understand its core assumptions, (3) be able to test the validity of these assumptions, (4) be able to test the robustness of findings produced by the method, (5) understand the method’s sample size requirements, and (6) understand the limits (real and imagined) to the generalizability of the method’s findings. Toward this end, the workshop will provide an overview of the method, its theory and its practice. Each issue addressed will be illustrated by examples with a focus on why, when and how to use regression discontinuity analysis for education research.

Design and Analysis of Clustered Data
Spyros Konstantopoulos & Kimberly Maier, Michigan State University
This workshop will discuss the practical issues that arise when educational researchers design studies and analyze data with clustered structures. We will take an applied approach to the introduction of these ideas, linking the research questions to the most suitable randomized design and will demonstrate the utility of analytical methods that best model clustered data. The workshop will draw from several educational research examples.

Using Instrumental Variables In Education Research
Sean Reardon, Stanford University
Under the right circumstances, instrumental variables (IV) methods may provide unbiased estimates of the effect of a treatment when treatment status is not randomly assigned. As a result, IV methods may be very useful in education research. Nonetheless, IV methods rely on a number of relatively strong assumptions and yield estimates that may generalize to only subsamples of the population of interest. This workshop will provide an overview of the basic logic and assumptions underlying instrumental variables methods, emphasizing a conceptual understanding of IV methods, their usefulness, and their limitations.
In particular, the workshop will address the following topics:
1) the conceptual logic of IV methods
2) the conditions required for IV methods to be valid
3) the interpretation of IV estimates
4) sources of bias in IV estimates
5) the use of IV methods within the context of randomized experiments and multisite randomized experiments.

Designing Adequately Powered Cluster Randomized Trials
Jessaca Spybrook, Western Michigan University
The purpose of this workshop is to help researchers learn how to plan adequately powered cluster randomized trials. The workshop will focus on two-level cluster randomized trials, three-level cluster randomized trials, and three-level multi-site (or blocked) cluster randomized trials. In addition to providing the rationale and statistical framework for calculating the power of a study, participants will learn how to use the Optimal Design Software (a free program), a user-friendly program for calculating the power for various cluster randomized designs. The workshop will combine lecture components with hands-on practice utilizing the Optimal Design Software. Participants should bring a laptop to the session.

spacer
spacer
Copyright SREE
Site Map Privacy PolicyContact Us