The U.S. News and World Report “Best Colleges” rankings, which will be published next month, are viewed as a Baedeker and Bible by more than 5 million American parents considering colleges and universities for their high-school juniors and seniors.

We think that parents should use this guide with caution.

Our problem with the rankings begins with how they are produced. An unspecified number of academic leaders -- presidents, chancellors, provosts -- are sent questionnaires and asked to rate their peers. Though few of those surveyed have sat in on freshman lectures at Harvard or Yale (where students famously complain about the quality of undergraduate instruction), they invariably give the brand-name schools their highest endorsements.

Interestingly, when another college guide, the Princeton Review, surveyed students, half of the top 10 schools that U.S. News lauded for having “a strong commitment to undergraduate teaching” received a “C” grade or less.

More disturbing is that many of the measurements “Best Colleges” uses are subject to manipulation. As the survey has grown in importance, some school administrators have found ways to game the ratings race. One example: U.S. News ranks a school, in part, by the number of applications it receives compared with the number of students it accepts. Schools with higher rejection rates do better than those that admit a greater percentage of their applicants. We know of a young woman who was hired by the admissions office of a highly rated liberal-arts college to increase application numbers. The school wasn’t increasing its freshman slots, but was encouraging more teenagers to apply so that its rejection figures would be boosted.

Blurred Numbers

Acceptance rates, real or manufactured, are one of “16 indicators of academic excellence” the U.S. News editors consider as they grade the nation’s colleges and universities. Others are graduation rates, alumni donations and class sizes. The latter is unquestionably key to any student’s success. But here, too, U.S. News gives the schools room to blur the numbers.

Stanford University says 68 percent of its classes have fewer than 20 students. Strictly speaking, this is true. What isn’t reported is that these small classes are mainly highly specialized advanced courses such as “The History of Tobacco Use.”

Matter of Class

As per the U.S. News survey questions, Ohio State claims that 32 percent of its classes have fewer than 20 students. But this statistic won’t tell a student much. The way Ohio State comes up with this number is by rolling its many specialized seminars along with Psychology 101 -- which jams 623 students into a lecture hall -- into one reported statistic. A prospective student wants to know how many of those 600-head lecture courses exist and how many of them he or she will have to take; the way the survey is structured, it obscures that crucial figure.

The report considers “faculty resources,” meaning salary levels, as a major criterion in scoring. But six-figure pay for tenured professors doesn’t always translate into benefits for undergrads.

Often at highly ranked schools, the senior faculty members teach “one and two,” meaning three courses a year -- and those may well be graduate seminars. At wealthier research universities, senior faculty get a sabbatical every three years. That means a history major at Harvard may not find a mentor for a senior thesis because 42 percent of the history faculty, was - - last year -- off campus.

Measuring Ratios

Another questionable rating in the U.S. News survey is the student-faculty ratio. In it, the University of Southern California can proclaim that it has one professor for every nine of its students. What it omits, in this self-rendered report, is that the student pool includes doctoral candidates, who command most of the faculty’s time. Moreover, the faculty numerator adds in adjuncts and other part-timers, who make up more than half of USC’s instructional staff. Students who pay $41,022 a year expect more than an adjunct, no matter how talented or conscientious they might be.

Odder yet, the University of Colorado tells U.S. News that an eye-catching 86 percent of its teachers are full-time. This is because the U.S. News formula allows the institution to permute its 959 part-timers into 319 “full-time equivalents,” which enhances the faculty part of the ratio. Parents genuinely need to know if their son or daughter is going to be taught by adjuncts, graduate assistants, college seniors or computer programs. At the end of the day, these loopholes in “Best Colleges” don’t just obscure the facts; they distort them.

Helpful Parts

That said, some aspects of the “Best Colleges” report could be helpful. One useful figure we found was a table on how many students at a given school paid the full tuition tab. At Duke, 59 percent pay its $40,472 sticker price, as do 60 percent at Penn ($40,514) and 67 percent at Washington & Lee ($40,387). Even selective colleges need cash, and despite claims that they admit in a “needs-blind” fashion, they are more likely to take you if your parents don’t request a discount. If you’re not affluent or a potential scholarship student, you might be wasting your time by bothering to apply at these schools.

Though it doesn’t specifically survey for it, U.S. News also offers a kind of left-handed index of student satisfaction. Colleges report how many of their graduates contribute in a typical year. As we read “Best Colleges,” we were struck by how many grads don’t send their alma mater even a symbolic check. At Harvard, 63 percent give nothing at all. At Vanderbilt, 74 percent demur, as do 87 percent at UCLA. The non-donors are probably pleased to have degrees, but could quite a few have less than fond memories of their campus years?

Or perhaps they feel that they already “gave” when they paid sky-high tuition bills?

(Andrew Hacker and Claudia Dreifus are the authors of “Higher Education,” which was published in paperback this summer. The opinions expressed are their own.)

To contact the writers of this article: Claudia Dreifus at claudreif@aol.com

To contact the editor responsible for this article: James Greiff at jgreiff@bloomberg.net