Whodunit?

CONTINUED

It was Stratfield's track record on the Iowa Tests of Basic Skills that triggered the tampering investigation. The Iowas were administered to district students in January, but before sending the exams out for scoring, two central office administrators picked over the answer sheets of the 153 Stratfield 3rd and 5th graders who took the test. Students from the school do extremely well on the Iowas--from 1990 to 1992, composite scores never dipped below the 98th percentile--and the administrators were hoping to find clues as to what other schools might do to score higher. What they found, though, was an unusually high number of erasures. The tests were forwarded to the Riverside Publishing Co., the Chicago-area publisher of the Iowas, where researchers confirmed district officials' suspicions: Not only did the Stratfield tests contain an unusually high number of erasures, but the percentage of answers changed from wrong to right on at least one subsection was also high.

More study was needed. District officials retested students at Stratfield and two other schools on March 21 and forwarded the results to Riverside. On April 15, the publisher's researchers delivered their report. Stratfield scores on the March retest had dropped on some sections by as much as 10 points. Perhaps more troubling was Riverside's now-completed study of erasures on the January tests. The number of Stratfield erasures was higher--five times higher in some cases--than the number of erasures on other schools' tests. Also, the researchers found an unusually high number of erasures--89 percent--that changed the answer from wrong to right. At the other schools, the highest correction rate posted by any...