The decline in SAT verbal scores

Magnifying glass over the stack of booksBetween 1963 and 1979 SAT-verbal scores declined 50 points. This is often attributed to change in the demographics of test-takers, a much broader segment of the population now takes the SAT tests. However, Cornell University researchers Donald P. Hayes, Loreen T. Wolfer and Michael F. Wolfe point out inconsistencies in that explanation, and suggest that simplification of school textbooks lies behind the decline in student’s reading comprehension and verbal achievement.

They describe a simple, low-cost experiment that school districts can carry out themselves to test how textbook difficulty affects their student’s verbal achievement levels.

Hayes et al. point out three discrepancies between students’ performances and the explanation that it is due to the changing demographics of students taking the test. First, scores did not drop during the 1950-1963 period when the population taking the test changed the most and grew the fastest.

Second, scores should have leveled off when the growth and composition of test takers leveled off (remaining fairly stable at 50 percent of the high school population between 1963 and 1979), yet this was precisely when the big decline in scores occurred. Third, the number of students scoring over 600 should have remained about the same.

Instead the entire distribution of verbal scores, from top to bottom, shifted to lower levels. There are now 35 percent fewer students scoring over 600. The number scoring over 700 dropped from 17,500 in 1972 to just over 10,000 in 1993, even as the number taking the test grew. Highly selective colleges report mean verbal declines of about 40 points.

Simplified textbooks may be to blame

Hayes et al. suggest a different reason for the decline, which was first described by Jeanne Chall in 1977. A change in educational philosophy after World War II encouraged publishers and educators to make schoolwork more accessible by reducing the pace of instruction, lowering student workloads, assigning fewer written papers, and simplifying textbooks.

When publishers simplified textbooks, they reduced the depth and breadth of academic knowledge covered, and in doing so, limited the number of uncommon words in texts. After 11 years1 exposure to simplified textbooks, a cumulative deficit in students1 knowledge and advanced verbal skills is likely, according to these researchers.

This hypothesis of cumulative deficit implies a time lag between the production of simplified texts after World War II and a significant effect on students1 learning. The majority of new textbook series did not appear until after 1950; most schools continued to use the harder, older series until the mid-1950s. Children therefore had been exposed to the simpler readers for several years by the time the SAT-verbal scores began to drop in 1963.

Text simplified after WWI and WWII

These researchers analyzed more than 800 texts from all major and most minor publishers, using between 10 and 30 pages of each text (at least 1000 words) as a sample. The most difficult readers generally were published before 1918. After World War I and again after World War II, reading levels were simplified. Today’s mean sixth-, seventh- and eighth-grade readers are simpler than fifth-grade readers were before World War II.

Sentences also were shortened from about 20 words to about 14 words for grades 4-8. Since this period of simplification, however, the difficulty of primary readers has been restored to pre-World War II levels, while middle and high school readers remain much simpler than before World War II. Science textbooks have remained about the same over the years.

While acknowledging that changing demographics probably accounts for a portion of the downward trend, Hayes et al. conclude that simplifying textbooks has had an unexpected and undesirable effect on student1s knowledge and skills that is reflected in the dramatic decline in SAT-verbal scores. This hypothesis predicts that unless there is a significant increase in the reading level of schoolbooks, achievement will continue at this low level for many years.

An Experiment

These researchers propose a simple and inexpensive experiment to test their hypothesis. They call for large school districts to match pairs of same-grade classes in different schools. The districts must be large enough to detect the small accumulating effects of comparing large numbers of matching classes using less-difficult and more-difficult texts over a four-year period. If the hypothesis is proven correct, then schools would be in a position to insist that publishers raise the levels of new textbook series.

Schoolbook Simplification and Its Relation to the Decline in SAT-Verbal Scores, American Educational Research Journal, Volume 33, Number 2, Summer 1996 pp. 489-508.

Published in ERN November/December 1996 Volume 9 Number 5

Leave a Reply

  • (will not be published)