When the Education Department announced a decline in the student loan default rate last month, officials did not rush to declare a victory in the long campaign to control what has grown into a $2 billion problem, saying they needed to study data on 5,266 schools more carefully.

Their caution turned out to be justified. A student loan agency in California has discovered and reported to the Education Department a statistical error that means the default rate was virtually unchanged.

Besides obscuring a lack of progress on the default problem, the error artificially reduced the high default rates of some trade schools and community colleges, making their students again eligible for one type of federally guaranteed loan. In that particular program, called Supplemental Student Loans, defaults were growing rapidly before Congress last year excluded schools with default rates above 30 percent.

Education Department officials have acknowledged the mistake although they insist on calling it "an anomaly." They said the department was deciding whether or how to recalculate the default rates and would decide if affected schools are eligible for the supplemental loans on a case-by-case basis. The officials have refused to assign or accept responsibility for the error.

"I prefer not to put the fault on anybody. It's not anybody's fault," said Roberta Dunn, deputy assistant secretary for student financial assistance.

The Office of Postsecondary Education computes default rates for the nation and individual colleges by dividing the number of newly delinquent loans by the number of loans that became due in a fiscal year.

The error in the 1988 rate involved loans totaling almost $400 million that were "suspended" because the bookkeeping of a California loan processing company was insufficient to determine their status right away. According to the California agency, the suspended loans numbered 183,510.

The 1988 rate of 15.6 percent that the Education Department announced counted all of these loans as coming due, but none of them as being in default. If the loans were omitted from both categories, the national default rate would be about 17 percent, nearly the same as the 17.3 percent reported for fiscal 1987.

Department officials previously said most of the reported decline of 1.7 percentage points was attributable to technical changes that omitted schools with fewer than 30 borrowers, others that no longer participate in federal aid programs and foreign colleges where U.S. students are enrolled.

Dunn conceded it was "improper" to count the suspended loans in only one category, but would not characterize the announced default rate as inaccurate. She used the word "anomaly" because "suspended" is not a designation ordinarily used in accounting for student loans.

The error was discovered by the California Student Aid Commission, a state agency that approves federal guarantees for student loans made by banks. Samuel M. Kipp III, the commission's executive director, said employees noticed that default rates at a number of California schools had curiously dropped below 30 percent. They tracked the changes back to the suspended loans.

"These were inadvertently counted as being in repayment," Kipp said. He suggested the mistake was made by a government contractor that compiled data from computer tapes submitted by guarantee agencies such as his.

Kipp said a further review identified 23 California schools whose students appear to be eligible for supplemental loans, but should not be because their correct default rates exceed 30 percent. He estimated that at least 100 schools nationwide were in a similar position.

Another commission official, Greg Gollihur, said the California agency has not guaranteed any new supplemental loans for students at the 23 schools, but added it was possible that other agencies around the country have approved loans involving other ineligible schools.

Dunn said the department was unsure how to resolve the issue because of the potential for legal challenges from schools that would be adversely affected if a recalculation were done. "We don't know what we can do," Dunn said.

The General Accounting Office has reported that supplemental loan defaults grew from $14 million in fiscal 1987 to $247 million two years later. The supplemental loans, which have higher loan limits than guaranteed student loans, were designed primarily for graduate students, but came to be used mostly by trade school students. A congressional aide said preliminary evidence indicates that the rapid growth in supplemental loans has slowed since Congress imposed restrictions in the budget reconciliation bill last year.