How to manipulate student test results
How to manipulate student test results
Most states have figured out that the fastest way to improve student performance on federally mandated standardized tests is to make the test easier or to lower the passing grade. States regularly trumpet their success in meeting the demands of the No Child Left Behind Act as the percentage of students deemed “proficient” increases every year. Since each state determines what is proficient, this standard is of dubious value. Not surprisingly, the National Assessment of Educational Progress, over which the states have no control, is not finding the rates of improvement that many states are claiming.
In an op-ed in the July 25, 2006 Wall Street Journal, Charles Murray of the American Enterprise Institute describes a less obvious way in which the state test results are misleading. Many states are also reporting success in closing the test score gap between black and white students, but Murray explains in some detail how a little statistical sleight of hand combined with less demanding standards makes this possible. Murray shows that with current test scores one can make the case that the gap is closing by lowering the standards or that it is growing by raising the standards.
One need not buy Murray’s analysis of test scores that he presented a few years ago in The Bell Curve to see that he has a useful point to make about how some states are playing fast and loose with test results to paint the picture they want. The real picture is that student performance is not what it should be and that the gap between black and white students remains disturbingly large. Pretending that we’re making progress will not fix the problem.
For those who can’t get enough of statistics and favor nature in the nature/nurture debate, Murray recommends the website of La Griffe du Lion (www.lagriffedulion.f2s.com). The site includes extensive analysis of test scores as well as discussions of why a Kenyan from the Nandi district is more than 500 times as likely as a European to be a top 1,500-meter runner. The website will not win a politically correct seal of approval, and the controversial nature of its opinions/findings probably explain the La Griffe’s use of a nom de plume.
Most states have figured out that the fastest way to improve student performance on federally mandated standardized tests is to make the test easier or to lower the passing grade. States regularly trumpet their success in meeting the demands of the No Child Left Behind Act as the percentage of students deemed “proficient” increases every year. Since each state determines what is proficient, this standard is of dubious value. Not surprisingly, the National Assessment of Educational Progress, over which the states have no control, is not finding the rates of improvement that many states are claiming.
In an op-ed in the July 25, 2006 Wall Street Journal, Charles Murray of the American Enterprise Institute describes a less obvious way in which the state test results are misleading. Many states are also reporting success in closing the test score gap between black and white students, but Murray explains in some detail how a little statistical sleight of hand combined with less demanding standards makes this possible. Murray shows that with current test scores one can make the case that the gap is closing by lowering the standards or that it is growing by raising the standards.
One need not buy Murray’s analysis of test scores that he presented a few years ago in The Bell Curve to see that he has a useful point to make about how some states are playing fast and loose with test results to paint the picture they want. The real picture is that student performance is not what it should be and that the gap between black and white students remains disturbingly large. Pretending that we’re making progress will not fix the problem.
For those who can’t get enough of statistics and favor nature in the nature/nurture debate, Murray recommends the website of La Griffe du Lion (www.lagriffedulion.f2s.com). The site includes extensive analysis of test scores as well as discussions of why a Kenyan from the Nandi district is more than 500 times as likely as a European to be a top 1,500-meter runner. The website will not win a politically correct seal of approval, and the controversial nature of its opinions/findings probably explain the La Griffe’s use of a nom de plume.
posted by Kevin Finneran at
5:00 PM
0 Comments:
Post a Comment
<< Home