There I was, parent of a second-grader, looking at test scores used by Fairfax County public schools to evaluate pupils for selection to the gifted and talented program in third grade. The more I stared at the scores -- and at FCPS's explanation -- the more questions I had. What did the numbers mean? What part would they play in the GT selection process? What else would the selection process involve?
This is the story of my effort to find enough dots to form a coherent picture of how FCPS selects students for the [full-time] GT program.
The first piece of information a parent sees in the GT selection process is a set of scores on the Cognitive Abilities Test. CogAT results are shown as percentiles and as standard age scores. The cover letter with the scores explains that "a percentile of 45 would mean that the student performed as well as or better than 45 percent of the students in the national sample." But what is a standard age score? The letter tells only the range and midpoint of standard age scores. What information does the standard age score offer that the percentile score does not?
I found the answer at a Web site of Riverside Publishing Co. [www.riverpub.com/products/group/cogat6/short.pdf], which publishes the CogAT. A document on the site explains that percentile and standard age scores are just two different measures of the same thing, with the standard age scores allowing finer distinctions at the extremes.
But how are the scores used?
At a meeting in February, FCPS representatives talked to parents about tentative plans to include in their "initial screening pool," from which GT students are selected, the top 10 percent of scorers on each of the three tests considered: two iterations of the CogAT, plus the Naglieri Nonverbal Ability Test. But the percentile scores reported to parents, an FCPS representative explained, show only how children score relative to the rest of the nation; Fairfax County children score higher on average. She warned me that my daughter's 97th percentile score might not make the 10 percent cut; to be safe, she said, I needed to go through the alternative procedure, in which parents refer their children.
In my efforts to assess my daughter's position, I could find no one at FCPS who could tell me how to relate national percentiles to Fairfax County percentiles -- they didn't have the score data broken out that way.
Wondering how hard that analysis could be, I tried it myself. Using the Microsoft Works database software that came with my computer, I needed only a few seconds to sort a database of 10,000 records, representing FCPS's 10,000 or so second-graders, into numerical order. From there, it was easy to find the top 10 percent: it's just the top 1,000 records. Why couldn't FCPS come up with that kind of breakdown?
Parents of the children selected to the initial screening pool were notified. Parents of the children whose scores didn't make the cut were not notified unless they had referred a child under the alternative procedure. They had to learn through the grapevine that they didn't make it.
For those children selected to the screening pool, files were forwarded to the Central Selection Committee, which, in most years, selects about 5 percent of second-graders to enter the full-time program. According to FCPS's "GT Center Screening Procedures," the committee "looks for compelling evidence that a child's needs cannot be met in a general education classroom." Beyond a list of documents the committee reviews and FCPS's statement that the committee uses "a holistic, case study approach," I could not find an explanation of just how they make that determination.
Although the people at FCPS seem to be trying hard to respond to parents -- they certainly answered my endless questions with great patience, for instance -- parents need and deserve access to better information at every step in the GT selection process.
One idea might be to show test scores in all formats -- standard age scores, national percentile and countywide percentile -- on one graph so that parents can see how the three relate. For interested parents, FCPS could add to its Web site a brief statistical explanation of the scoring or a link to Riverside's plain-English explanation of test scores. And FCPS could use its Web site to keep parents better informed of things like evaluation criteria: By what standard does FCPS decide who is selected for the screening pool and, finally, for the program?
As parents and as taxpayers, we have a right to know how our children are being evaluated and classified.
Why can't FCPS just tell us?
Hacker lives in Reston.