TRANSPARENCY TUESDAY: No need for your No. 2 pencil. Just bring your skepticism.
Two recent announcements have spotlighted the importance of public records in bringing to light what now appears to be rampant dishonesty among school employees in fudging standardized test scores to improve their schools' rankings.
Item One was the news from Oklahoma that the state was throwing out test results from six schools where investigators concluded scores were artificially inflated. The state's information came from tipsters, from classroom observations, and from spot-checks of the test papers themselves; a pattern of similar erasures in which wrong answers are changed to correct ones is a red flag for tampering.
Item Two was an investigative reporting award recognizing the collaborative work of a team of journalists led by USA Today, who produced a multi-part 2011 series that suggested standardized-test cheating was commonplace, and rarely punished.
The journalists' research -- covering six states and the District of Columbia -- found more than 1,600 "statistical anomalies" in which test scores showed improbably large improvement year-over-year.
Investigative Reporters and Editors, Inc., honored the series, "Testing the System," with second prize in the annual Philip Meyer Journalism Awards, which recognize innovate use of social science research methods in journalism.
The USA Today-led team built its findings on a "standard deviation" analysis looking at the statistical likelihood of drastic swings in student performance from one year to the next. In many hundreds of instances, the rate of improvement was so far beyond what history has shown to be likely -- without the aid of cheating -- that the scores were flagged as suspect.
That is some fairly high-level statistical analysis, and not easily replicated. But there now is so much attention being paid to suspicious gains in testing performance that journalists can, at least, find out whether their state Department of Education or school district has conducted an "erasure study" or a statistical analysis of testing performance patterns. The findings of such investigations, as well as any punitive or corrective actions taken in response to the findings, should be a matter of public record in all states.
As with many instances in which journalists seek records from schools, it is likely that some institutions will wave the flag of "student confidentiality" in an attempt to hide behind the Family Educational Rights and Privacy Act.
Don't let them.
FERPA privacy applies only to students' individual test papers or test scores, and most certainly not to statistics. So long as a particular score is not matchable to a named student, aggregated testing data is not FERPA information, and it should be readily obtainable by way of an open records request.Tagged: public records, Transparency Tuesday