|Having trouble viewing this email? Click here.|
In This Issue:
Synthetic Data Provides New Tool for Researchers
Researchers interested in accessing Texas student administrative data come to The University of Texas at Dallas Education Research Center (UTD-ERC). The UTD-ERC's abundant data warehouse, rich with student data, higher education data, and more, is an unparalleled resource for researchers across the country interested in delving into important education issues.
Only approved researchers, though, may access the data which means, because of privacy issues, researchers cannot view data until their proposal has been approved. So, how do they develop an understanding of the data? Data documentation provides some level of understanding. Another tool, newly developed by the UTD-ERC, is synthetic data.
The UTD-ERC offers over 500 million individual education data records, which fall into over 40 categories and date back to the early 1990's. With this first version of synthetic data, data file structure has been replicated so that researchers can now test code before using the real data.
Synthetic data is available for each data category via the Documentation page of the ERC website, texasschoolsproject.org. These flat file data can be imported into the appropriate SAS, SPSS, or STATA format and used to develop and test programs before proposals are developed.
Because this is the first version of such data, comments and suggestions for improvement are welcome and can be directed to Jim Parsons.
Texas Schools Project continues its regular 2009-2010 Seminar Series with a presentation by Dr. Scott Imberman, University of Houston:
Wednesday, February 3
Using panel data from a large urban school district, Dr. Imberman provides the first arguably causal evidence on the impact of school uniforms on student outcomes.
For more information about Dr. Imberman and his presentation, go to the Events page of our website.
For future planning, these seminars occur on the first Wednesday and third Friday of each month.
There has been much discussion about teacher quality and the effect of such on student achievement. Dr. Steve Rivkin, economist and professor at Amherst College, presented findings from his study with Eric Hanushek, "Do Disadvantaged Urban Schools Lose Their Best Teachers?" at a November Texas Schools Project seminar.
In the study, Rivkin and Hanushek use data from a large urban district in Texas to investigate teacher quality differences by transition status and school characteristics.
During the seminar Dr. Rivkin remarked, "Perhaps the more important result . . . is that teachers who switch schools within a district, switch districts, or exit the Texas public schools entirely do not appear more effective on average than those who remain." In fact, their study finds that teachers who leave the most disadvantaged schools are less effective on average than those who stay.
The full paper can be viewed on the Events page of our website under Dr. Rivkin's November 13 seminar listing.
As all researchers know, when testing a treatment, experimental random assignment is the gold standard. Random assignment, however, isn't always possible - especially in education. To test treatments when random assignment hasn't occurred, regression discontinuity design (or RD design) can be a solid evalution method.
With RD Design, participants are assigned to a program or comparison group solely on the basis of a cutoff score (cut point) from a pre-program measure. Causal inferences can then be made for those close to the cut point.
Dr. Brian McCall, a University of Michigan econometrician well-known in the field of RD design, shared information on this quasi-experimental evaluation approach during the November 20 Texas Schools Project seminar.
"While RD design has been around for over 50 years, it was dormant for quite some time," said McCall. "It has seen a resurgance of use in the last ten years, though, and has been used recently in education research to study areas such as financial aid, class size, test scores, and the effects of remediation on college outcomes."
Why such a resurgance? McCall explained, "Under fairly general conditions, most of which are testable, RD design allows researchers to make causal statements regarding the impact of a treatment on an outcome."
Critical observations using RD design are those which occur around the cut point. Dr. McCall remarked, "We get something very close to randomized assignment just around the cut point. People just above look very similar to those just below."
During the seminar, Dr. McCall also discussed the importance of satisfying various assumptions for the design to be valid and ways to check assumptions using various techniques.
He was followed by Dr. Steve DesJardins, also of the University of Michigan, who, with Dr. McCall, applied RD design to the evaluation of two educational programs, the Gates Millennium Scholars program and the Washington State Achievers program, programs focused on improving college and workplace readiness for outstanding minority students with significant financial need.
To view the papers, slides and videos of Dr. McCall and Dr. DesJardins' presentations, visit the November 20 seminar listing on our Events page.
Advanced Placement (AP), because of its continued growth and student exam pass rates, has recently been a news topic, both locally and nationally. And each time, reporters have turned to Dr. Kristin Klopfenstein, a senior researcher with Texas Schools Project, for her insight.
Dr. Klopfenstein was asked to participate in a New York Times article on Advanced Placement in December.
In the article, Klopfenstein commented, "The original point of the AP program was to make college-level study possible for advanced high school students . . . But, the AP program has been transformed to serve many more purposes. For students, taking AP courses signals academic ability and work ethic to prospective colleges. For high schools, having a lot of AP classes signals quality to the community and real estate markets. For educational reformers, offering the program has become a way to provide academic rigor with accountability in the form of standardized end-of-course exams. All of these forces are jointly responsible for the dramatic growth in the AP program over the last 15 years . . . These newer functions have introduced substantial challenges and distortions."
A December 6 Dallas Morning News article on the number of Texas students failing AP exams said, ". . . the latest data show Texas high school students fail more than half of the college-level exams, and their performance trails national averages."
In the article, Dr. Klopfenstein commented, "The disappointing scores show that educators and policymakers may have misused Advanced Placement in their zeal to ready students for college." She added that it raises the question of whether resources spent on AP programs might better invested other areas. And, as is pointed out in the article, "That may already be happening, as some high school students are instead opting for dual-credit classes because they provide college credit without having to pass the AP test."
(Read more) of The New York Times article, "The Advanced Placement Juggernaut."
(Read more) of The Dallas Morning News article, "More Texas students taking, failing Advanced Placement exams."
Magnus Lofstrom is a research fellow at the Public Policy Institute of California. He is also a research fellow with the Institute for the Study of Labor (IZA), a research associate with The Center for Urban Economics, and a research affiliate with Texas Schools Project. Prior to joining PPIC, Dr. Lofstrom was an assistant professor of economics at The University of Texas at Dallas.
Areas of interest for Dr. Lofstrom include labor economics, economics of immigration, economics of education, and public policy. His work has been published in many scholarly journals including Journal of Human Resources and The Future of Children.
Recently, Dr. Lofstrom presented, with Dr. John Tyler, a study entitled, "Is the GED an Effective Route to Postsecondary Education for School Dropouts?" at the American Economic Association's Annual Meeting.
Using data from Texas Schools Project, Lofstrom and Tyler examined the extent to which dropouts use the GED as a route to post-secondary education, basing their estimates on a set of academically at-risk students who were very similar in the 8th grade.
In the study, they observed that the eventual high school graduates in this at-risk group have much better postsecondary education outcomes than do the similar at-risk 8th graders who dropped out and obtained a GED. This recent study adds to Dr. Lofstrom's extensive research in the areas of GED and dropouts.
"Is the GED an Effective Route to Postsecondary Education for School Dropouts?" (Read more)
For more information on other papers written by Dr. Lofstrom on GED and dropouts, visit the Research Areas section of our website.
The University of Texas at Dallas Education Research Center (UTD-ERC), part of Texas Schools Project, houses a wealth of data provided by the Texas Education Agency, Texas Higher Education Coordinating Board, and State Board for Educator Certification.
A complete listing of current data available can be found on our website under Data Holdings.
Information on accessing this data can be found on our website under Access.
To unsubscribe from this monthly e-newsletter, email [email protected] with "unsubscribe tsp-news" as the subject.