In issuing its report last week on trends in reading performance since 1971, the National Assessment of Educational Progress once again omitted results from its “anomalous” 1985-86 test.
Instead, it released a new analysis of the findings, which confirmed that the anomaly in the test’s results was most likely caused by a change in the test design, and did not reflect a significant drop in student achievement.
When the 1986 results were first compiled, in 1987, naep officials noted large unexplained declines in the performance of 9- and 17-year-olds, compared with the 1984 results. (See Education Week, Jan. 20, 1988.)
Because these results appeared er4roneous, and did not appear to reflect genuine declines in student achievement, naep officials agreed to delay reporting them until after naep and an Education Department study group had completed separate studies. The report on the results, released in 1988, did not compare the findings with those of prior years.
In addition to studying possible causes, the officials agreed to test students during the 1988 assessment to determine if the 1986 results were genuine.
A report on that analysis, “The Effect of Changes in the National Assessment: Disentangling the naep 1985-86 Reading Anomaly,” concludes that average reading proficiency declined slightly between81984 and 1986 at all age levels, and that in 1988 students rebounded to their 1984 levels.
But these results “are not fully understood,” the report states, and naep officials again agreed not to report the “questionable” 1986 results.
The report concludes that the declines observed in 1987 probably reflect the fact that the assessors asked 1984 questions in a different context in 1986.
In the future, the authors recommend, naep should use new items to gauge long-range trends only after they have been used in two consecutive assessments.
“When measuring change, do not change the measure,” the report concludes.
