Married American women who usually say they prefer an equal number of sons and daughters, appear to have an underlying bias for sons and only a weak commitment to a sexually balanced family, according to a study released today.

The study was published in Family Planning Perspective, journal of the Alan Guttmacher Institute, an affiliate of the Planned Parenthood Federation of America Inc.

According to the study, results from a survey of 6,800 married women of childbearing age indicate that 49 per cent had an underlying preference for boys.

Only 32 per cent said they would like to have daughters. About one-fifth (19 per cent) expressed a preference for an equal number of each sex.

The studys author, Lolagene C. Coombs of the University of Michigan's Population Studies Center, used a new first preferences and underlying bias, according to the institute.

Underlying biases could be important, Coombs said, because they might help predict how far couples would go to achieve the sexual ratio they want between sons and daughters.

Although most of the preferences found in each group were "mind" rather than "strong," Coombs reportedly found nearly twice as many strong preferences for boys as for girls, by 20 to 11 per cent.

Still, Coombs said, the preferences observed in the study may not affect overall fertility. "Limited available data suggest that most couples would not continue to bear children beyond the number they want in order to have the sex ratio desired."

The study is based on interviews conducted in 1973 in a National Survey of Family Growth sponsored by the Health, Education and Welfare Department's National Center for Health Statistics.

In writing the report of the study, Coombs said clues to the roots of these sex preferences must come from beyond that collected for the 1973 study.

"Sex preferences for boys are often explained by their value as adults (to carry on the family name or for support in old age, for example) . . .," wrote Coombs.

She also said that preferences for girls are usually based on their companionship and providing help around the house for mothers.

Coombs asked if these implied that women who have strong son preferences are more future-oriented, while those with daughter preferences are more concerned with immediate needs. "There is no evidence on this point," she replied to her own question.

In concluding her article, Coombs wrote: "We suspect that sex preferences are more deeply rooted in the culture and reflect a constellation of cultural attitudes about sex roles and values. The implication of sex preferences for family composition are probably much broader than their importance for fertility behavior, great as that may be in some parts of the world."