A U.S. Justice Department conclusion that District whites are victimized by violent crimes at twice the rate of District blacks may not be so much at odds with the D.C. police department's figures that show blacks are more likely to be victims of reported violence, according to criminal justice experts and statisticians.

Blacks appear to be the likely victims in the most serious of the violent crimes, which also are more likely to be reported to police, while whites appear to be more likely victims of less serious crimes of violence that may often go unreported, they said.

"When the seriousness of the offense increases, then the difference between the races decreases," data in the Justice survey indicate, said Patrick A. Langan, a Bureau of Justice Statistics statistician who worked on the survey. "I think there is a heck of a correspondence between what we found and what they the police find."

One of the survey's conclusions was that District whites are victimized by all types of violent crimes the surveyors asked about at a rate of 110.2 per 1,000 white residents over the age of 12, while blacks are victimized at a rate of 57.4 per 1,000 black residents.

The crimes included robbery, assault (including rape) and threat to injure. Homicides were not included because the survey relied on victim interviews.

Police officials had said earlier that the most serious crimes come to their attention and that their figures show higher victimization rates for blacks, except for robbery.

A police department theory that a telephone survey would be more likely to pick up minor crimes that go unreported to police is "very reasonable," Langan said.

National statistics show that a higher percentage of blacks are the victims of "aggravated assault," attacks involving a weapon or requiring at least two days of hospitalization, while more whites are victims of "simple" assault without a weapon or serious injury, Langan said. The Justice Department survey grouped both types of assaults together in its analysis; police department rates include only aggravated assault.

U.S. Attorney Joseph E. diGenova said the survey's results did not indicate any need for a change in law enforcement policies approaches but did show a need for more reliable information about crime.

"It means we get different types of information depending on how we look at this problem," diGenova said. "I was surprised by their conclusions on the racial differences . I won't reject them just because I was surprised by them. I want to look more closely at them."

He said he had no explanation for the higher white victimization rate and agreed with the police department that most of the victims in the District that his office sees are black. "We know from what we see in the courts that there is a majority of intraracial crime."

The type of victim survey used for the Justice Department report is a widely used and methodologically sound way of gathering crime data, according to criminal justice experts.

Albert D. Biderman, a criminal justice expert and statistician with the Bureau of Social Science Research Inc., who was involved in the survey, said the study "went about it the right way" and that the figures are "probably as accurate as those that measure employment or other economic statistics, in some ways better."

But Biderman also pointed to the large amount of threatened rather than actual violence included in the findings.

While some threats, such as constant bullying at school or spouse intimidation, may be as serious if not more so than actual injury, "it is different than the image we have of violent crime," he said. "You find higher rates of victimization because you get many more things are not stereotypical [violent crimes]."