Dr. Alice Stewart looked carefully at the high-powered criticism of her December study linking cancer to low-level radiation, took the study back to the laboratory and returned yesterday with the same results: workers at the government's nuclear facilities in Hanford, Wash., died of cancer at a rate 5 percent greater than the general population.

Stewart's presentation to the American Association for the Advancement of Science convention was the latest salvo in the running battle over what, if anything, should be done about low-level radiation.

"Now, please don't say I like radiation," pleaded one of Stewart's critics, Seymour Jablon of the National Academy of Sciences, who directs the medical followup agency of the academy's National Research Council.

"Certainly there's something peculiar going on among the Hanford workers," he said, but noted that it may be chemical exposures and not radiation. "Let's not rule out other possibilities," Jablon added.

Stewart's initial study was begun in 1975 by Dr. Thomas Mancuso of the University of Pittsburgh, who subsequently lost his job at the old Energy Research and Development Administration, which had tried to delay publication of his results.

He published anyway, with Stewart and Dr. George Kneale as coauthors, in December. The Department of Energy has ordered a study of why Mancuso lost his job, and a spokesman said yesterday a report is expected in about three weeks.

"The real reason why so many experts have found fault with our analysis is because our estimates of risk are much higher than earlier ones based on A-bomb survivors," Stewart said in the written revision.

With Mancuso and Kneale listed this time as her coauthors, Stewart wrote that their findings are supported by new studies of some of the 40,000 persons who entered Hiroshima and Nagasaki to look for relatives or possessions within four days of the 1945 blasts. Previous studies focused on actual bomb blast survivors.

Like workers at Hanford, Hiroshima returnees received low-level doses of radiation, although the Hanford workers received similar doses over a much longer time. There is substantial controversy over whether the body "repairs" damage done by low radiation between doese, or whether the total dosage over years must be counted when assessing risk. Stewart said her study still leaves that question open.

"We know that a small dose only has a small effect, but if it has that effect and we can measure what it is, then we know we must do something about setting appropriate standards for exposure," she said. She noted that the average worker under study received a radiation dose over five years equal only to that received all at once by a person taking a barium enema, a procedure commonly used to diagnose diseases of the lower intestine.

Such a procedure is well below current safety limits. Those limits, Stewart said, probably ought to be reexamined in light of her findings. "It isn't safe at any dose," she said.

Lauriston Taylor of the National Council for Radiation Protection and Measurements, a privately funded research and consulting organization, disagreed. "If we act prematurely, we may deprive outselves of uses of radiation that could be very important to our way of living," he said.

"This work is interesting and importnat but not yet definitive . . . compared with the risks (to workers) in most other industries, this one is better than most. There's no such thing as totally safe and unsafe in life."

Jablon pointed out that further research on worker histories at other nuclear laboratories and at nuclear power plants is needed but is hampered by laws surrounding the public disclosures of causes of death. "If those bombs (on Japan) had been dropped in the United States, we still wouldn't know what had happened to the survivors," he said.