Rock stars, movie actresses and heads of state have shined a bright light on global poverty in the past year, often highlighting the particular burden on women, but a report from the United Nations released this week painstakingly details the huge gaps in data needed to understand how poverty - in all its ugly guises - affects women.

Many poor countries simply do not collect the most basic facts about births, marriages and deaths by sex and age, or the employment status and wages of men and women. 

The dearth of information makes it difficult to pinpoint where girls are being married off while they are still children, or where female fetuses are being aborted because boys are preferred, or where girls are dying because they get less food and medical care than boys, says the report, which was released Wednesday.

Its authors, and specialists in the field, say better information is urgently needed if the world is to fashion sensible, effective solutions to reflect conditions that are constantly evolving and vary greatly even within a single country.

Africa, the world's poorest region, has the weakest systems for data collection. Four in 10 Africans live in countries that did not conduct a census in the past decade, and 8 in 10 live in countries with inadequate national collection of vital statistics.

India and China, home to half of the world's poor, carried out censuses but have weak systems for registering births and deaths, information needed to understand trends in health and sex discrimination at the local level, said Mary Chamie, who heads the demographic and social statistics branch of the United Nations statistics division.

"If we don't know what the problem is, how are we going to fix it?" she said. "And if the situation is changing, how will we know there's been a shift if we plan based on information from a decade ago?"

The myriad reports published each year about poverty in developing countries often include caveats about flaws in the data, but such warnings are often over shadowed by sweeping conclusions. The new report, "The World's Women, 2005: Progress in Statistics," is the first to comprehensively analyze which countries collect official statistics by sex and which do not, its authors say.

Jody Heymann, director of the Institute for Health and Social Policy at McGill University in Montreal, has spent a decade studying working families globally. She said she had found that many developing countries did not collect data about hours worked, benefits provided or sex gaps in wages. "And there's practically nothing on who's caring for the children," she said.

Statistics about the diseases that cause deaths in African adults are inadequate, said Angus Deaton, an economics professor at Princeton University. Household surveys can help determine how many babies and children have died, because their parents usually survive to tell about it. But when parents die, families often splinter and no one is home when a surveyor knocks.

International organizations like the United Nations that are charged with assembling data about how many Africans die of causes related to AIDS "have only scattered clues," Professor Deaton said. As to how many African women die in pregnancy or childbirth, he said, "That's a complete black hole, because we don't know about adult mortality."

A lot has been written about the missing women of South Asia and China - the female fetuses aborted or the girls who died prematurely because of sex discrimination. Attention has also been paid to worsening life expectancies for men in Eastern Europe and Russia. But better data are needed to address many such problems, the report concludes.