A student's academic achievement is partly determined by his teachers' preconceptions of his ability, a Harvard lecturer believes.

Robert Rosenthal of the Social Relations Dept. discovered four years ago that rats performed far better when he told experimenters, falsely, that the rats had been specially bred for intelligence.

Children for Rats

Rosenthal then substituted teachers for experimenters and school children for rats. He got much the some results.

Rosenthal took a random sampling of first-and-second-grade children at a South San Francisco elementary school, and told teachers that these students would make dramatic gains in school work. His randomly chosen group made those gains, while the rest of the student body did not. Only the teachers--not the pupils or parents--had been told of the predictions.

Rosenthal noted that some professors "have been saying for a long time that some children are victims of educational self-fulfilling prophecies, but they just haven't been able to come up with data to prove it. We think we have."

Rosenthal began his test in the spring of 1964. With the permission of the principal of San Francisco's Spruce School, where students are divided into three "tracks"--fast, medium and slow--he gave an IQ test to all students in the school's kindergarten and first grades.

Teachers were falsely told that the test would show which pupils were due to "spurt ahead" academically. The teachers were given the names of 20 percent of the student body, randomly selected from all grades and all three tracks, and were told that every pupil so listed would improve dramatically within a year.

A year later, when all the children still in school were re-tested, the spurters showed an average I.Q. gain of 12.22 points, compared with 8.42 for a control group representing the rest of the student body.

The most dramatic gains came in the first and second grades, where the "spurters" increased 27.4 (in grade one) and 16.5 (in grade two). The control group rose only 12 points in the first grade and 7 in the second.

The Spruce School has a large Mexican enrollment, and this provided an interesting sidelight. Among the "spurter" Mexican boys with faces that loked Anglo-Saxon showed higher gains than those with more identifiably Mexican faces.

"There is no clear explanation for this finding," said Rosenthal, "but we can speculate that the teachers' pre-experimental expectancies of the more Mexican-looking boys' intellectual performances were probably slowest of all."

No Corroboration

Rosenthal has conducted similar experiments at elementary schools in a Massachusetts suburb and a middleclass town in Ohio. Although Analysis of these two tests has not been completed, the results have not clearly corroborated the Spruce School findings."

"They show that teacher expectations ca nbe a very powerful determinant," he said. "What they don't show is that if teachers expect more, they will necessarily get more."