In a vigorous defense of his report on public and private schools, sociologist James S. Coleman said here yesterday that stronger academic demands and better student behavior were the key reasons why private schools generally produce higher academic achievement than public schools -- not who controls them or who goes to them.

Among public schools, Coleman said, the ones with more homework, less absenteeism, and an orderly environment had higher achievement regardless of the family background of their students, which is the same phenomenon he said he found more frequently in private schools.

"The research has begun to accumulate that there are some characteristics of the schools themselves that school policy can do something about that do lead to better achievement," Coleman told a conference at the L'Enfant Plaza Hotel about his massive, government-sponsored report.

"Public schools are subject to more constraints than they were 20 years ago," Coleman continued. "But we must learn how to carry out good education in orderly environments, and I think we can."

Coleman's new report, made public last week, contended that Catholic and other private high schools provide a better education than public schools do and in some respects are less racially segregated. The report was financed by the National Center for Education Statistics, a part of the federal Education Department.

In earlier statements and at yesterday's conference, which NCES sponsored, several researchers challenged Coleman's statistical procedures for comparing public and private schools and for measuring segregation. But even among critics admiration was expressed for the massive scope of Coleman's descriptive data.

"It's remarkable," said Donald Erickson, an expert on private schools at the University of San Francisco. "Coleman's predictions don't tell us what will happen. They create a whole lot of static . . . but the data is unprecedented. For the first time we will be able to get some evidence on important issues. . . . I have a feeling that Coleman has done it again."

Now 54, Coleman is a professor at the University of Chicago and a principal investigator for the National Opinion Research Center, which has the NCES research contract. In 1966, while at Johns Hopkins University, Coleman headed a massive study of educational opportunity for blacks and whites that was widely used to support school desegregation. Later, Coleman said further research on how desegregation was carried out did not support its conclusion that black achievement would increase.

The new report includes data on 58,728 students in 1,016 schools, who were tested and surveyed last spring. It is the largest U.S. study of private and public schools ever conducted.

Data in the report show that among public school students three-quarters say they do less than one hour of homework a night, only 6 percent take three years of a foreign language, and 46 percent report that "students often don't attend school." In Catholic and other private schools only about half the students reported less than an hour's homework a night. A third year of a foreign language was taken by 14 percent of Catholic high students and 20 percent in other private schools. Serious attendance problems were reported by only 8 percent in Catholic schools and 16 percent in other private schools.

Coleman said other recent studies, particularly one of government-run schools in London by Michael Rutter, had also shown higher achievement in rigorous, orderly schools. Some of his critics at the conference countered that the greater selectivity of private schools and their ability to expel poorly behaved students were the underlying reason for the different climate and results. Coleman and his defenders said public schools could adopt policies to achieve the same success.

An internal National Center for Education Statistics memo, which was given to reporters several days before the conference, criticized the way Coleman handled some of his data. But at yesterday's meeting, which drew about 400 scholars, lobbyists and administrators, the center distributed an elaborate rebuttal in which Coleman confirmed his main point that while private schools have more affluent students than public ones, similar students of all family backgrounds do better in private schools.

In a separate paper, the Rev. Andrew Greeley, a University of Arizona sociologist, said his analysis of the data shows that students from poor black and Hispanic families with lesser academic skills show the most improvement in Catholic schools.

"It's the people at the lower end of all the hierarchies who are getting the greatest payoff from the Catholic schools," said Greeley, a Roman Catholic priest who has often been at odds with church leaders because of his reports on the widespread use of contraceptives by American Catholics."These findings are the opposite of what I expected and of what the [Catholic] schools themselves think they are doing."