After a four-month examination of freshman admissions, University of California officials released data Monday that didn't yield an absolute answer as to whether race still plays a role.

The admissions probe was ordered last fall after John Moores, chairman of the UC Board of Regents, issued a report that found hundreds of students with low SAT scores were admitted to Berkeley in fall 2002, while more than 3,000 with stellar scores were not.

His report also found that more than half the low-scoring students were African American or Latino.

In response to Moores' criticism that UC was rejecting bright students in favor of mediocre ones, UC President Robert Dynes appointed a group of UC administrators, faculty and students in October to scrutinize the university's 3-year-old comprehensive review process.

The group will release its full findings to the Board of Regents next week, but shared new data Monday that don't eliminate the possibility that race and ethnicity still play a part in admissions decisions. Voter-approved Proposition 209 banned UC from using affirmative action in admissions starting in 1998.

In its complex analysis, UC found that in 2003, African Americans and Latinos were admitted at higher rates than predicted at its six selective campuses - Berkeley, UCLA, Davis, San Diego, Irvine and Santa Barbara. Fewer Asian American students were admitted than predicted at those campuses, except UCLA.

"We need to analyze whether it's part of the statistical noise in our model or whether unintentional preferences are being exercised on some campuses," said Bruce Darling, senior vice president of university affairs and co-chairman of the study group.

UC used a complicated model that accounted for grades, SAT scores, income level and other quantifiable factors to predict admission. Prediction rates did not include intangible admissions criteria, such as leadership potential and overcoming personal hardship.

At UC Berkeley for example, the analysis predicted that 234 African American students would have been admitted in 2003, while 355 actually were. By contrast, 4,433 Asian American students were predicted to have been given an offer to Berkeley; 4,214 were. At UC Davis, 1,953 Latino students were predicted to have been admitted; 2,020 were.

At Berkeley and UCLA, the number of white students who received admission offers fell short of predictions.

Those differences however, are much smaller compared to data from 1997, the last year UC could use race. That year, for example, UCLA admitted 456 African American students, nearly triple the 160 predicted by the model.

Still, UC officials contend the discrepancies in 2003 are less troubling than the overall numbers that show that African American and Latino students continue to be underrepresented at the university.

UC officials also stopped short of saying that their data invalidate Moores' report, which only looked at SAT scores.

"I think (the public) will have to judge that for themselves," said Richard Black,assistant vice chancellor for admissions and enrollment at UC Berkeley. "We have never been impressed that Moores' unidimensional approach has validity."

Moores has suggested that comprehensive review - which looks at nonacademic factors such as hardship, talent and leadership abilities - may have skewed Berkeley admissions in favor of students from low-performing high schools, in which African American and Latino students tend to predominate.