Though Numbers May Be Accurate, Some Data Hold Very Little Meaning
The New Jersey education department last month released its annual school report card, a compilation of information on every school in the state.
And to help parents compare schools, The New York Times published a table of some of the information, including test scores, per-pupil funding, and graduation information. Among other data, the table said that Camden's Woodrow Wilson High School had a graduation rate of 123 percent, one of the highest in New Jersey.
Could this be the same school that writer Jonathan Kozol highlighted in Savage Inequalities, his 1991 book on the plight of urban school systems? Mr. Kozol wrote: "Of the nearly 1,400 children who attend this school, more than 800 drop out in the course of four years. About 200 finally graduate each year."
In fact, the school report card numbers may very well be accurate. But they are not very meaningful, contends Philip Burch, a Rutgers University research professor who has spent several years studying New Jersey's dropout and graduation rates.
The report card, which uses the term "graduation data" rather than "graduation rate," includes this definition of the figure: "Number of students graduating with class by August 1996 as percent of grade 12 enrollment in October 1995."
In other words, the figure doesn't include students who dropped out before the 12th grade. And students who transferred into a given school--or were simply not in school for the beginning-of-the-year head count--can inflate the figure, even above 100 percent. In fact, several other urban high schools also had percentages well above 100 percent.
But, said Mr. Burch, to be meaningful, the rate should consider more than just the senior year.
"Using the 12th grade is a ridiculous way of approaching the subject of graduation rates," he said, "because most of the dropouts occur, especially for the minority students, before the 12th grade, so you miss all the prior dropouts."
To calculate school and districtwide graduation rates, Mr. Burch uses the ratio of graduates from each school to its 9th grade enrollment when those graduates were freshmen. This ratio is called a cohort rate because it seeks to track a group of students over the four years of high school.
Based on this approach, Woodrow Wilson High School has a rate closer to 50 percent.
This is how the state education department computes the statewide graduation rate, as well as the graduation rate for each of its 21 counties in an annual report called "Vital Education Statistics." But by publishing the graduation rate this way only by county, Mr. Burch believes the state's report glosses over the plight of New Jersey's urban districts.
For instance, the vital statistics show Camden's countywide graduation rate in the 1994-95 school year was 82.5 percent--not far from the state average of 83.5 percent. But this masks the fact that the rate in the Camden city district was far lower, Mr. Burch said.
State officials in New Jersey say they present the best information they have in both the state report card and in the statistics reports.
Computing a true cohort graduation rate for each school or district is made impossible by the great transience among students in many districts, which makes it difficult to distinguish students who drop out from a school from those who transfer.
"We don't have the mechanism for tracking the students," said Gary Reece, the assistant state education commissioner who oversees the report card project. Student populations in such urban schools as Woodrow Wilson are particularly transient, he said.
Officials at the U.S. Department of Education's National Center for Education Statistics say the inability to account for student mobility hampers many states in attempts to calculate a true graduation or dropout rate. But a few states have resolved the problem.
The Texas Education Agency, for example, tracks students by their Social Security numbers or some other identification numbers. That way, the state can remove students from a school's dropout rolls if their identification numbers show up at another school or if they earn a General Educational Development certificate.
To address concerns over confidentiality, the state agency instituted safeguards to restrict agency personnel from matching student names with their identification numbers and other records.
New Jersey began moving toward such a program in the last school year by assigning numbers to students when they take the state's standardized tests.
In the meantime, New Jersey includes in its school report cards the enrollments in each grade level over the past four years.
"That was our attempt to tell the truth," Mr. Reece said.
At Woodrow Wilson High, for example, the report shows that in the 1992-93 school year, it had 399 freshman. The next year it had 276 sophomores. The year after that it had 242 juniors, and in the 1995-96 school year it had 175 seniors.