Better student scores challenge old beliefs RESEARCH REPORT

January 08, 1995|By M. William Salganik | M. William Salganik,Sun Staff Writer

Here's news: Student achievement is up in the United States. Minority achievement is up more than the national average. And changes in family structure are contributing to the increase.

A study released last month by the RAND Corp. challenges conventional wisdom about American education, throwing data against the view that achievement is declining and that family pathology is a factor in the decline.

"There is a perception that deterioration has occurred," in school achievement said David W. Grissmer, a senior analyst at RAND and study leader. "The data show that has not happened."

Called "Student Achievement and the Changing American Family," the study takes data from two federal studies that followed students over a number of years.

Using statistical analysis techniques, it estimates the effect on achievement of family factors. A higher education level of parents, for example, leads to higher achievement scores by students. Having a single-parent household reduces scores, on average.

The net effect of family changes over the past 20 years, the study found, is positive -- so scores should have increased. That's because of two major shifts in family structure:

* Increases in parent education. Mothers of 15- to 18-year-old students were more than twice as likely to have a college degree (15.8 percent vs. 6.7 percent) in 1990 as they were in 1970.

* Smaller family size. In 1970, 48.2 percent of students lived in households with one or no siblings. By 1990, that was up to 72.5 percent. Since average family income, adjusted for inflation, changed little over the two decades, this means more money was available per child.

Working in the opposite direction, there were more births to teen-age mothers (13.1 percent of 1990 students had been born to mothers under 19, compared with 9.3 percent in 1970). However, the researchers found that the negative effect this caused on achievement was smaller than the positive effect from the other trends.

There was also an increase in mothers working outside the home (54.0 percent in 1970, 67.5 percent in 1990), but the study concluded that working mothers had a little effect either way on achievement scores.

rTC And in fact, scores did increase on the National Assessment of Educational Progress (NAEP) tests, which the authors call "the best data to monitor the achievement trends of U.S. students over the last 25 years."

Scores went up slightly for white students -- 3 percentile points -- but jumped 11 points for Hispanics and 19 for blacks.

The white increase was about what would have been predicted from changes in families, such as increased parent education levels. "There were no dramatic changes in educational productivity outside of minority students," Mr. Grissmer said.

The black and Hispanic gains were much larger than predictions, leading the authors to conclude that desegregation efforts, increased school spending or other policy changes had actually resulted in improvement.

But the study doesn't say which of those factors had a positive impact. Next, Mr. Grissmer said, the RAND researchers hope to look at changes regionally to try to see what helped. "Per pupil spending was up everywhere," he said, "while integration occurred rapidly [only] in the South." So, region-by-region scores might show which policy changes mattered most.

Also, the researchers hope to do more studies of the so-called "multiple-risk" students -- those who are, for example, from low-income families and with low parent-education levels.

In these follow-up studies, Mr. Grissmer said, he hopes to "put the whole question in an investment framework." Health researchers, he said, have done much better modeling than education researchers into cost-effectiveness of preventive measures, such as inoculation.

If achievement is actually up over the last 20 years, why are so many Americans convinced it is dropping?

Part of the problem, Mr. Grissmer said, is a concentration on the Scholastic Assessment Test, a test commonly taken for college entrance. Scores on the SAT are down over the last 20 years. But since students decide themselves whether to take the SAT, the study says, the "constantly changing proportion and composition" of students taking the test make it a poor measure of changes over time.

Much better, the researchers said, is NAEP, which is taken by a representative sample of students.

Further, Mr. Grissmer hypothesized, a bleak picture of social conditions is often created by "advocacy statistics" in which people concerned about a problem select and publicize facts to highlight their concerns.

"This is endemic to democracy," he said. "When we see a problem, groups organize, trying to make a case and get attention. It happens on both the liberal and conservative sides. We often generate a picture of ourselves that's at odds with what's happening on the average."

The problem is exacerbated, he said, by education statistics that are not nearly as definitive as in other areas. "No one," he said, "would argue about whether the gross national product has gone up or down."

Baltimore Sun Articles
|
|
|
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.