(The New York Times, 05 November 2006)

ISABELLE CARBONELL, a college senior from Bethesda, Md., has thrived over the last four years as part of a small learning community. Most of her classes have had fewer than 35 students. For freshman and sophomore years, her dormitory was in the same building as the cafeteria and many of her classrooms and professors’ offices. ”You see the same people over and over, and that lets you create networks,” she says. ”You get to know your professors informally. You see them in the hallway, they say, ‘How’s that project going along?’ and you bounce ideas off them.” Prospective undergraduates are deluged with statistics — from average class size to the number of Nobel Prize winners on staff — with which to take the measure of a college. Is Ms. Carbonell’s story an argument for choosing your college by the numbers?

Not exactly. She attends the University of Michigan at Ann Arbor, which has some 25,000 undergraduates, 4,100 full-time faculty members and 540 buildings. If she had been looking for an intimate experience, the numbers would have led her elsewhere. That’s just one of the problems with statistics: they rarely tell the whole story. (Ms. Carbonell’s story is that she signed up for Michigan’s Residential College, a program in which students live and attend classes in the same building. She now lives off campus but continues to take classes in the R.C. building.)

Another problem with numbers: ”Often statistics don’t measure what’s important,” says Lloyd Thacker, executive director of the Education Conservancy, a nonprofit group working to improve the college admissions process. For example, the selectivity of a college, measured by how many applicants it denies, provides little information about the educational experience there.

Also, statistics can be fudged. Regard any number you read in a glossy brochure or on a university Web site with skepticism, says Arthur Levine, president of the Woodrow Wilson National Fellowship Foundation. ”It’s not information,” he says. ”It’s marketing.” This fall Margaret Spellings, the secretary of education, announced a plan to push colleges to clarify and expand the data they report about themselves, including collecting information about student learning.

Until that comes to pass, though, the college- bound will just have to make do. Here are six common statistics that students in search of a quality college should parse carefully.


You want most of your classes to have about 25 students, says Marty O’Connell, a spokeswoman for Colleges That Change Lives, a nonprofit organization that promotes the ideas espoused in Loren Pope’s books of the same name. ”These are settings where students can be asked to be critical, creative thinkers and to collaborate with their peers and teachers,” she says. ”Faculty members cannot get to know students if you have 300 in a room.”

But don’t depend on a college’s averageclass- size statistic to predict the density of your classroom. As an average, it distorts reality by including, say, the tiny Tagalog seminars you’ll never take with the popular politics courses you will.

The same can be said of student/faculty ratio: a campus can have a 13-to-1 ratio and you can still end up in a class of 50.

That’s because not all professors teach undergraduates — and some don’t teach at all. Rebecca Goldin is director of research at Statistical Assessment Service, a media watchdog group affiliated with George Mason University. She notes that it’s a common practice in large science departments for professors to use grant money they raise for research to pay back a percentage of their salary to the university, with the intention of getting out of teaching one or more of their courses. The university then turns around and hires an adjunct to fill in at a lower salary and counts both individuals as faculty members.

Instead of looking at the student/faculty ratio, ask how often you can expect to have small classes: every semester or just during your last two years? ”It should be all the time, on a regular basis,” says Ms. O’Connell.

The statistics typically used to indicate quality of instruction are the number of Ph.D.’s on staff, the number of Nobel Prize winners, the number of National Academy of Science members, and faculty salaries. Ignore them all, says Marty Nemko, an independent career and college counselor in Oakland, Calif., and the author of ”The All-in-One College Guide.”

”People who self-select into Ph.D. programs are academic research types, not teachers,” he says. ”Their knowledge is so deep and so profound they often don’t have the ability to communicate well with undergraduates who need the basics. In addition, they get their status not through the quality of their teaching but the quality of their research.”

He adds: ”A person with a Nobel Prize-winner mind is in the loftiest stratospheres of their arcane pursuit and, in general, is not that gifted a teacher. And measuring teaching quality by salaries is ridiculous. Teachers are paid more when they’re bringing in more research dollars and when the college is in an area with a high cost of living.”

What’s a better quantitative measure of teaching quality? The percentage of fulltime faculty members at a college, says Ms. O’Connell. Full-timers should be in the majority, she says, because being around tends to increase their participation in the life of the campus and their students’ development.

Look at a college’s retention rate — the percentage of students who come back for their sophomore year — to gauge student satisfaction, suggests Ann Wallace , director of counseling and guidance services at Rye Neck High School in Mamaroneck, N.Y. ”I don’t like anything below 80 percent,” she says. ”Students should have questions if it’s low. Was it financial aid that didn’t get followed up with the second year? Or was it dissatisfaction or location? There can be reasons to explain it, but it’s generally not a good thing.”

To determine whether you’ll fit in with the students, scan average SAT scores, says Dr. Nemko. First, decide whether you want to be among your intellectual peers, hang out with people smarter than you, or be a big fish in a less-selective pond. ”Once you’ve made that determination,” he says, ”as much maligned as the SAT is, it’s quite a valid indicator of intellectual firepower and drive.”

One caution: Colleges have been known to exclude the lowest scores when computing averages, and calling those students ”special admits.” ”Everybody lies about their college boards,” says Dr. Levine, former president of Teachers College at Columbia. ”They keep combing through, and they get rid of some of their most troubled populations.”

This statistical sleight of hand can easily increase an average SAT score. In its online ”Facts at a Glance,” the College of New Jersey, for example, reports an SAT math and reading average of 1300 for ”regular admits”; for all entering freshmen it was actually 1255.

The problem, Dr. Nemko says, is you don’t know which college is tweaking and by how much. ”The way you get around it is asking the point-blank question: ‘When you include all students who are in the freshman class, what is the average SATscore?’ ”

Diversity statistics can also help students figure out whether a college is a good match for them. Chris Farmer , college counselor at the Young Women’s Leadership School in East Harlem , says: ”Sometimes a college will say ’35 percent of our students are students of color,’ and the kids at my school will say, ‘Wow, I can see myself there.’ But in their minds they’re thinking that 35 percent of students are African- American or Latino when most of them are Asian.” So ask for a breakdown.

He suggests translating percentages into numbers, especially for smaller, less diverse colleges. ”Say you have a college with 500 freshmen,” he says. ”If they have 3 percent African-American kids, that means about 15 in the freshman class are black. Because many more African-American women go to college than men, you might want to look at that. Maybe only one or two of them are men.”

Mr. Farmer cautions that statistics for students who get jobs within a year of graduation tend to be inflated because ”the students who report back to the college are the ones who are really satisfied with their job situation.” A better, though more limited, indication of a college’s ability to prepare students, he argues, is a high percentage of students who are accepted at graduate schools. But be on the lookout for spin.

”The college might say ’80 percent of our students who applied to medical school got accepted,’ ” he says. ”The real answer might be that, in junior year, there were 15 pre-med majors who are no longer pre-med majors because their G.P.A. was too low. So although their intention was to go to medical school, the college counseled them to not apply.”

Simply graduating from college is an achievement — only 55 percent of students at four-year institutions do — so you should seek out colleges where advancement is the norm. But graduation rates measure only the percentage of freshmen who earn their degrees from the colleges they began in; they don’t count the large number of students who transfer in and out.

Colleges promote their six-year graduation rates rather than the lower four-year rates, though that might be in small print. ”If a college is graduating 75 percent of students in four or five years, it’s probably pretty solid,” says Mr. Farmer. ”If they’re graduating 90 percent, it’s outstanding. When a college is graduating 50 percent of the students who start, that doesn’t mean don’t go there. It just means figure out why the students are leaving.”

Not all reasons are bad. Small liberal arts colleges may lose students who decide to pursue a field in which it doesn’t offer courses. Lower graduation rates in a rigorous program might indicate that its standards are high, Professor Goldin notes. ”If you want a challenge, it may not be the wrong decision to go there,” she says. ”But go into it with your eyes open.”

Read this article at the New York Times.