But all the critics agreed that school districts should brace for another deluge of paperwork when the Education Department concludes that another universal survey is necessary.
In 1984, when the Education Department’s office for civil rights altered the way school districts are selected for its semiannual survey, education and civil-rights groups argued that the changes would severely reduce its usefulness.
In planning for the 1988 survey--which provides the only federal data on civil-rights compliance in education--O.C.R. has quietly inched back toward its old method, rescinding a change that allowed large districts to sample only certain schools and designing the sample to include more districts that have not been surveyed recently.
But an Education Week analysis of agency records indicates that, even with the planned changes, the 1988 survey will again bypass many districts not surveyed recently. That group includes more than half of all school districts in the country.
O.C.R.'s failure to survey all or most districts periodically is the chief complaint of its critics, who argued in 1984 that the agency should retain a system that ensured that all districts above a certain size would be surveyed at least once every six years.
About 2,000 districts that were included at least once in the 1978-1980-1982 cycle will have been bypassed by the surveys done in 1980, 1982, and 1988, according to the analysis.
In addition, about 7,000 districts, mostly small ones, will not have been included since 1976, when virtually every district in the country was surveyed.
The Education Week analysis tabulated the number of districts included in each of the past six survey years for 11 representative states, which encompass almost 6,000 school districts. It also tabulated the number of districts in those states slated for inclusion in 1988.
The results were then extrapolated to estimate the number of districts included nationwide. According to agency documents, the sampling rate does not vary widely from state to state.
The analysis utilized OCR’s 1986 estimate of the “universe’’ of all school districts: 15,800.
O.C.R. officials said the agency does not keep track of the number of districts that are not surveyed, and refused to compute the figure.
Gary Curran, an agency spokesman, said the number of bypassed districts is “irrelevant.’'
“The system we use is the system we need,’' Mr. Curran said. “It allows us to pick up 75 percent of the minority population.’'
But OCR monitors compliance with laws forbidding discrimination based on gender and disability, as well as on race.
The purpose of the survey, which collects data on the racial makeup of student populations, assignment to gifted and special-education classifications, disciplinary actions, and other such subjects, is to bring to OCR’s attention districts whose statistics are suspect for some reason. The agency can then review those districts for civil-rights compliance.
The survey is designed to include most districts that are of obvious interest, such as large, urban districts and those where problems have surfaced in the past.
But critics argue that a sampling method that does not survey all or most districts periodically will not only miss individual problem districts, but will eventually erode the database, making it impossible to project national trends accurately. (See Education Week, Sept. 12, 1984.)
The erosion, critics say, would also render inaccurate those sampling techniques designed to select districts that are nationally representative, and could force OCR to do another “universal’’ survey like the 1976 effort.
Likening OCR officials to “civil-rights cops,’' Paul Smith, research director at the Children’s Defense Fund, said: “You can’t tell which neighborhoods haven’t had a prowl car go through for four months.’'
History of Controversy
The school civil-rights survey was first conducted in 1968, and was completed annually until 1975. Sampling methods varied, but aimed to include most large districts and to yield a representative sample from which national statistics would be projected.
The strategy changed in 1976, partially because ‘s predecessor agency within the Department of Health, Education, and Welfare had been charged with enforcing laws prohibiting discrimination on the basis of gender and disability that were enacted in 1972 and 1973.
The 1976 survey included every school in the nation, and requested more data than had ever been required before.
The political controversy over the survey was so intense that the Office of Management and Budget decided to cancel it, and reversed course only under strong pressure from civil-rights groups.
The reaction from school officials was equally furious. That year, a Massachusetts school superintendent wrote: “I have just received from the Office for Civil Rights 129 pages, in sextuplicate, of bureaucratic balderdash. ... In our opinion, these reports are unrealistic, wasteful of energy, time, and money, and unnecessary.’'
Likely To Recur?
OCR’s critics fear that same problem will recur if too many districts are repeatedly left out of the survey.
After 1976, the survey was made semiannual. The sampling method used in 1978, 1980, and 1982 was designed to include all “high-interest’’ districts; it was also to survey one-third of all other districts in each sample year, ensuring that all districts were surveyed at least once in a six-year cycle.
Small districts, however, were excluded. The threshold enrollment level was 300 students in 1978 and 1980, and 1,500 in 1982.
The threshold was eliminated in 1984, and some small districts have been included in subsequent surveys.
In 1988, according to Education Week’s analysis, some 1,600 such districts will be included in the survey.
But, because of the previous exemption, more than 7,000 districts will still not have been surveyed since 1976, according the analysis.
Other changes made in 1984 narrowed the scope of the sample in order to reduce the number of schools surveyed and thus the “paperwork burden.’'
The “rolling’’ sample, which had included a third of the districts each year, was abandoned in favor of a “stratified random sample,’' which intentionally includes high-interest districts and selects others at random, controlling for geography and size.
Additionally, large districts were given the option of “subsampling,’' or surveying only a portion of their schools.
Changes Were Opposed
Education and civil-rights groups, most notably the NAACP Legal Defense and Educational Fund and the Council of Chief State School Officers, vehemently opposed the changes. They argued against abandoning the rolling sample, and said subsampling would render data from those districts suspect.
“They never did give us any good reasons to abandon a perfectly good system,’' said Phyllis P. McClure, director of the legal defense fund’s Washington office.
“I think it’s a case where everybody meant well,’' said Mr. Smith of the Children’s Defense Fund. “They were trying to respond to the demand to reduce paperwork.’'
“But what this really meant,’' he said, “is that the burden would descend on small districts totally at random and completely by surprise. In the real world, what school districts value most is that whatever they have to do, they get plenty of advance notice, and it is perfectly predictable what they have to do.’'
1988 Techniques Differ
Mr. Curran of OCR argued that the key element is to survey districts where there are most likely to be problems, and that surprise can be helpful.
A rolling sample “telegraphs which school districts are going to be surveyed when, and as an enforcement agency, that’s not in our best interests,’' he said.
For the 1988 survey, OCR has dropped the “subsampling’’ option, which Mr. Curran said was used by only “about half a dozen’’ of 60 eligible districts.
The 1988 sample was also designed specifically to include more districts that had not been surveyed recently, Frank K. Krueger, acting director for policy and enforcement, said in a letter to Education Week.
The minor changes in the 1988 survey techniques do not solve the problem, OCR’s critics say.
‘Important for All Districts’
“It’s important for all districts to have to report this information periodically, or there’s no incentive for them to collect it at all,’' said Joan First, president of the National Coalition of Advocates for Students, an organization that disseminates survey data on individual districts and ranks states on various topics covered by the survey.
Ms. First said she fears her organization’s analyses will become increasingly inaccurate as the survey’s database erodes.
Mr. Smith is more concerned that O.C.R. will miss unsuspected trouble spots. He is less worried about national projections, he said, because the sampling is not controlled enough to make them accurate in any event.
But all the critics agreed that school districts should brace for another deluge of paperwork when the Education Department concludes that another universal survey is necessary.
