Andrew Miller grew up in southern California, attended a private high school, and earned the high grades that virtually ensured him a spot at nearby UCLA or USC. But a two-hour test in his junior year altered his path dramatically. Along with more than a million fellow high school juniors, Miller sat for the PSAT. Though typically described as practice for its better-known older sibling (the SAT), the test's lesser billing obscures its role as the lever over hundreds of millions of scholarship dollars in a national sweepstakes to attract the coveted top performers.
Miller's score of 215 (out of a possible 240) qualified him for consideration as a National Merit Finalist in 2003, putting him in the sights of schools around the country. He eventually picked the University of Oklahoma, which offers out-of-state National Merit Finalists the prize in financial aid: a "free ride" currently worth about $65,000 over four years.
OU is one of the nation's top contenders in the competition among schools to enrich their portfolios of National Merit Scholars. The current freshman class includes 170-more per capita than any U.S. public university. In addition to the more than $6.3 million that National CrossTalk estimates the university spends on scholarships and tuition waivers, OU courts this cadre of students with perks like special dorms and early class registration. "We make a big push to try to attract those students," said Associate Vice President Matt Hamilton.
It is a push that catches students' and parents' attention. Miller was entertaining offers from UCLA and USC when he learned about Oklahoma's arrangement. UCLA offers just $1,000 a year for National Merit Finalists without financial need, and even USC's $13,000-a-year offer wouldn't have covered his costs. "I wouldn't have been out here if it hadn't been for the scholarship," said Miller. "I love it here. You feel really special."
Indeed, for most involved in the National Merit Scholarship Program (NMSP), there is little to dislike. Students get seductive stipends. Schools get the status conferred by the trademark. Corporations get tax deductions and scholarship perks for employees. And the College Board, which co-sponsors the PSAT with the Illinois-based National Merit Scholar Corporation (NMSC), gets the prestige that ensures an expanding market.
Last fall, more than 1.4 million high school juniors entered the competition by sitting for the test officially known as the PSAT/NMSQT (Preliminary SAT/National Merit Scholarship Qualifying Test). For the first time, an even greater number of non-juniors took the exam, said Beth Robinson, executive director of the College Board's PSAT Program. Altogether, 2.94 million students took the test, yielding an estimated $32 million in revenue for the College Board.
The PSAT joined forces with the NMSP in 1971, and in recent years, the connection has escaped controversy. Despite a civil-rights complaint involving NMSP a decade ago, suspicions about race and gender score gaps have concentrated on the SAT, where an uneasy equilibrium has settled around cautions not to heed small score differences or rely on tests alone in making high-stakes decisions. The admonitions, however, seem to have bypassed financial aid offices and scholarship programs like NMSP, where small score differences on a single test can make all the difference.
|National Merit Scholar Andrew Miller received a four-year "full ride" scholarship to the University of Oklahoma.
(Photo by Lisa Hoke, Black Star, for CrossTalk)
Indeed, though National Merit Scholars are in the spotlight, the program itself operates in the shadows. But now, as it marks its 50th year, the NMSP's definition of merit has come into the crosshairs of some in academe. In March, a subcommittee of College Board trustees was examining the organization's connection to the NMSP. And a University of California faculty admissions committee, in a strongly worded letter, was urging campuses to cut ties to the program.
New scrutiny over the nation's premier merit scholarship could elevate questions about similar programs run by states and institutions, heighten the debate over merit- versus need-based aid, and exert pressure on the College Board, whose tax-exempt purpose includes a "commitment to excellence and equity."
To critics, the NMSP connection violates part of that commitment. "What intrigues me is how the College Board can get away with claiming it cares about equity and then endorsing and marketing this indefensible program," said Patrick Hayashi, a retired senior University of California official and former College Board trustee. "The College Board should not endorse any program that actually fosters inequality."
Hayashi's concerns, outlined in an eight-page letter last October, prompted the internal review. The letter asserts that during Hayashi's ten years overseeing admissions at UC Berkeley, not one of the hundreds of National Merit Scholars who came to the campus was black or Hispanic. College Board officials repeatedly turned down requests for an accounting of poor and minority students in the program, he wrote.
"I estimate that the percent of National Merit Scholars who are black, Hispanic, and American Indian is close to zero and that the absolute number of poor students from these groups is also close to zero," Hayashi wrote shortly before ending his term as a trustee. "If we ever learned the precise figures, then we would be forced to question the wisdom and morality of sponsoring a 'merit' scholarship program that effectively locks out black, Hispanic, and American Indian students...The College Board endorses it and gives it national reach and impact. Without us, the NMSP would not have nearly the prestige, acceptance and impact that it does."
Subcommittee members were guarded about their deliberations. Their information packet "was marked confidential in 20 different places," said Esther Hugo of Santa Monica College. Asked about NMSP, subcommittee chair Ted Spencer said, "I don't have any concerns about it." Spencer, director of admissions at the University of Michigan, said he could not discuss the committee's work until it was finished, but added, "I'm not certain we'll ever finish."
Another former trustee responded to Hayashi's allegations in an e-mail to fellow trustees last November. Bruce Walker, vice provost at the University of Texas at Austin, said, "I believe that this could be one of the most important issues we will have faced as trustees." While conceding that NMSP's processes are "antiquated," Walker suggested that the partnership could help increase diversity.
|Use of the PSAT to narrow the field of potential National Merit Scholars is acceptable to some testing experts, including Ed Haertel of Stanford University.
(Photo by Rod Searcey for CrossTalk)
"NMSC represents significant market advantages, and our partnership adds considerably to net revenues that can be used to accomplish goals for which there is no revenue stream...goals such as equity and access," Walker wrote. "I see no reason to sever our relationship with the NMSC. To do so would be an act of fiscal irresponsibility." In an interview, he declined to elaborate. UT-Austin is a top recruiter of National Merit Scholars, but Walker said NMSP is just one way of recognizing merit.
As the Trustees' subcommittee reviewed the partnership with NMSP, the nature of the contractual relationship remained murky. College Board's Beth Robinson mentioned royalties that support research, development, and the PSAT program. But NMSC spokeswoman Elaine Detweiler insisted that no money changes hands. This much is clear: The NMSC agrees to use the PSAT to select winners, and the College Board agrees to provide juniors' scores to NMSC.
Even before the Trustees' review was concluded, College Board representatives made their position clear. "The practice that National Merit is following is very consistent with the requirement that they use multiple sources of information in making a high-stakes decision," said Wayne Camara, vice president for research and development. A week after the interview with Camara, College Board trustees voted to continue the partnership for a year.
The use of multiple sources, avoidance of narrow score differences, and the caution to use tests only for purposes for which they were validated are hallmarks of professional guidelines established jointly by the American Educational Research Association, the American Psychological Association, and the National Council on Measurement in Education and of the Code of Fair Testing Practices, developed by testing officials, including Camara.
Whether National Merit meets those guidelines is subject to interpretation. The process begins in April, when NMSP staff identify the 50,000 juniors with the highest PSAT scores-eliminating more than 1.3 million others. In September, the list is reduced to 16,000 semifinalists, the highest-scoring students in each state.
NMSC refused to provide a list of state cutoff scores, though they are published on the Internet. They range from 202 in Arkansas and West Virginia to 222 in Massachusetts and Maryland, Detweiler confirmed. (To critics who question the "merit" of the MNSP, the state-by-state differences belie the program's claim to be "national.")
Consider what the cutoffs mean in Andrew Miller's case: With a 4.0 average at the University of Oklahoma and plans to double major in mechanical engineering and pre-med, the clean-cut 19-year-old is clearly a high achiever. But had he scored a single point lower on the PSAT, he would have lacked the National Merit pedigree and the free ride at OU. Had he been born a year later, he also would have missed the cut-that year, California students needed 216 points to qualify. No luck either if he had attended high school in New York, which had a 218 hurdle in 2003. Yet, according to Camara, the difference between a 215 and 216, or between a 215 and 218, is meaningless.
|Wayne Camara, College Board vice president for research and development, defends use of the PSAT in the National Merit Scholar selection process.
(Photo by College Board)
Only for the 16,000 semifinalists do factors other than tests enter the picture. In February, NMSP staff eliminate about 1,000 of them-those whose high schools do not endorse them, whose grades are not high, or whose SAT scores do not confirm their PSAT scores. The remaining 15,000 finalists are all eligible for National Merit Scholar awards. In a typical year, about 8,200 such awards are available. Colleges offer roughly 4,600, sending sums of $500 to $2,000 per student in the fall to NMSC, which distributes the funds back to campuses each semester.
Many schools offer thousands more directly to students. About 2,500 students receive one-time scholarships of $2,500 from NMSC, and around 1,100 corporate-sponsored scholarships worth anywhere from $500 to $10,000 a year go primarily to employees' children. Financial need is never considered.
While opponents stress that most students are weeded out by virtue of test scores alone, defenders note that other factors enter at the end of the process. Camara said winnowing students from 50,000 to 8,200 exceeds the "acceptable" ratio of 4:1. He added that since the PSAT accurately predicts SAT scores, it shares the SAT's statistical validity.
|The University of California faculty admissions committee, chaired by Michael Brown of UC Santa Barbara, has urged UC campuses not to participate in the National Merit Scholarship Program.
(Photo by Rod Searcey for CrossTalk)
The 11-member University of California faculty admissions panel found Camara's answers unconvincing. "A test cannot be validated by proxy," wrote the chairman, UC Santa Barbara education professor Michael Brown, to campuses in early March. "Students who fall but one point below the cutoff score are summarily eliminated from further review. In other words, the answer to a single question (which is well-within the range of psychometric error) can cause students to miss the cut." Others argue that 2:1 (16,000 semifinalists to 8,200 scholars) or even 16:15 (16,000 semifinalists to 15,000 finalists eligible for scholarships) is the operative ratio.
Testing experts differ on the alleged transgressions. To Ed Haertel, a Stanford education professor who helped write the Standards for Educational and Psychological Testing, the NMSP's workings are not troubling. "I don't see a grave injustice in the fact that you have to get a high score on the test in order to qualify, especially given that other things are also examined," he said.
Robert L. Linn, professor of education at the University of Colorado at Boulder, however, sees room for improvement. "It's better for the test not to be the sole factor. It would be better if somehow they could figure out how to put grades in it," said Linn, who has headed both the National Council on Measurement in Education and the National Research Council's Committee on Testing and Assessment.
"In principal, anything that has a single measure is faulty," agreed J. Michael Thompson, vice provost for admissions at USC. "But how do you do it otherwise?"
Such pragmatic concerns win the day in most discussions. "I would probably do the same thing if I were president of the College Board," said Linn. "I'd think it wasn't fiscally responsible to cut my revenue. It contributes to the volume of students taking the PSAT and how many times they take it. It would be good if National Merit did give demographic information about their scholars. You can anticipate that it wouldn't look good for them. It's pragmatism versus what might be more socially desirable."
College Board officials believe their goals squarely target the public good. More than a practice test, the PSAT helps schools improve teaching and increase college attendance among minorities, said Peter Negroni, vice president for K-12. "We acknowledge this as a problem," he said. "We as an organization are engaged in steps to try to do something about it."
But though more minorities are taking the tests, Negroni offered no evidence that racial score gaps are narrowing, much to the frustration of Hayashi and others at UC. "If we can never get the data, we'll never know where we are and where we hope to go," said Judy Sakaki, vice chancellor for student affairs at UC Davis, and a member of the NMSP subcommittee.
Last fall, the College Board accused the anti-testing group FairTest, of copyright infringement for posting average SAT scores by race, figures previously released by the College Board itself. The most recent data, for 2004, show whites scored an average of 1059, Asian Americans 1084, Mexican Americans 909, and African Americans 857. The averages for males and females were 1049 and 1005 respectively.
Tacit acknowledgment of scanty diversity in the NMSP came on the heels of the 1964 Civil Rights Act. With a 1964 Ford Foundation grant, NMSC established a parallel National Achievement Scholarship Program for Outstanding Negro Students (a name still used in its annual report). Though NMSC runs no programs for other minorities, the College Board started a National Hispanic Recognition Program in 1983.
"You could say that it's socially responsible," noted Brown. "But that's tantamount to admission of an adverse impact. It creates an ongoing impression that racial minorities just cannot compete with other individuals in a 'fair' and achievement-based system."
The College Board asks African Americans interested in the scholarship program to check a box on the PSAT answer sheet. Students are also asked to choose among eight racial categories. Still the NMSC and College Board claim not to know how many National Merit Scholars are minorities. "That is not relevant to the conduct of the National Merit program," said spokeswoman Detweiler.
"They (NMSC) have always had a reputation of being very very hard-shelled and very difficult to get information from and have a dialogue with," said Larry Gladieux, the College Board's former director for policy analysis. "They are sort of 'hear no evil, see no evil, speak no evil.' They take the position that these scholarships promote achievement and excellence. They don't get entangled in policy...and they don't want the College Board to, either."
A year ago, the NMSC took apparent action to address a racial policy development. In the wake of the U.S. Supreme Court's 2003 decision regarding college admissions, it quietly increased the number of National Achievement awards it sponsors from 450 to 700, because colleges stopped sponsoring them.
Still, the overwhelming majority of funds-$46 million out of $48.7 million-goes to National Merit Scholars. (That sum includes additional scholarships offered by corporations for PSAT high-scorers, but not the additional amounts that colleges like Oklahoma offer.) Rather than money, the Hispanic program provides, for a fee, names of the 3,300 highest-scoring students who achieve a minimum grade point average to schools "interested in communicating with prospective students of Hispanic heritage," according to the College Board's website. No College Board officials interviewed knew the origins of that program.
The last major challenge to the NMSC occurred in 1994, when FairTest, unable to obtain racial information on scholarship winners, calculated that 60 percent were male and, since girls get better grades in college, filed a gender discrimination complaint. The case was settled with the addition of a writing section to the PSAT. Females generally score higher than males on tests of writing skill.
It is no coincidence that the new assault originated in California, the first state to voluntarily adopt race-blind admissions. Nor is it surprising that Hayashi, once a chief aide to former UC President Richard C. Atkinson, is behind it. Atkinson earned national attention in 2001, when he advocated abandoning the SAT, calling the focus on tests "the educational equivalent of a nuclear arms race." The standoff led to the revised SAT that debuted this spring, with another writing assessment-this time an essay-settling the dispute.
Six out of eight UC undergraduate campuses, all but UC Berkeley and UC Riverside, still sponsor National Merit Scholars, with awards capped at $2,000 a year, officials said. Systemwide, UC invested $824,000 in 676 National Merit Scholars this year, far fewer than individual campuses elsewhere. The panel that unanimously recommended discarding the NMSP includes representatives from every campus, suggesting that several campuses will follow the committee's lead, in what may be the logical conclusion of the university's ongoing reevaluation of admissions practices.
"UC over the last four or five years has taken such a strong position, and basically led the effort to change the whole SAT test, that for us to continue supporting the NMSP, given the way they use the PSAT, I think is a problem," said Tom Lifka, assistant vice chancellor for student academic services at UCLA. This year, UCLA sponsored 84 National Merit freshmen and enrolled a total of 115, more than any other UC campus.
|Retired University of California official Patrick Hayashi calls use of the PSAT to select National Merit Scholars "indefensible."
(Photo by Rod Searcey for CrossTalk)
Current National Merit Scholars would not be affected by a change, and at least one is sympathetic to the concerns. "Say you had a cold or your cat died that morning and you did poorly on the test, that would preclude you from participating," said freshman Jonathan Beck at UC Davis. "Other scholarships look at more aspects of your high school career. Your GPA is not just one test. Your extracurricular activities are not just one night." Still, he said merit aid is essential for students like him, who can't receive need-based aid but also don't get help from their parents.
As similar considerations shape the debate over NMSP, clearly there is a residue of hostility toward standardized tests among college leaders, revealed by the standing ovation Atkinson received after his 2001 speech on the SAT. Whether a departure by UC will influence those leaders or the College Board to break with NMSP is less clear. The nation's most prestigious universities, Harvard first among them, recruit the students without dangling scholarships. When UC Berkeley dropped out three years ago, the campus fell from fourth to 30th place in the rankings published every February by the Chronicle of Higher Education.
But with college presidents eyeing these rankings, dozens of admissions officials jockey for position each year, a habit that may be hard to change. Based on information provided by universities, National CrossTalk conservatively estimates total award values of well over $100 million, with public flagships among the major players:
- This year, the University of Florida attracted 259 Merit Scholars, second only to Harvard. UF offers a four-year package of $24,000 to resident students, according to a university website. Non-residents receive $40,000 plus roughly $50,000 in tuition waivers. "We really go after those people aggressively, and that was a planned choice to try to upgrade our profile," Academic Advising Center Director Albert Matheny told the campus newspaper.
- At UT-Austin, which perennially ranks in the top five for National Merit enrollment, Texas residents receive a $13,000 package over four years. Non-residents receive $38,000, including $8,500 in waived tuition and $1,000 in scholarships annually. "We recruit them vigorously because they represent one type of merit, and it's a type that turns out to be important," said campus Vice Provost Bruce Walker.
- Arizona State University now ranks 12th, offering awards worth $50,000 over four years to state residents and $80,000 to out-of-staters. A press release lauding the "meteoric rise" from six National Merit freshmen in 1991 to 162 this year credits a recent increase in the students' stipend.
- The University of Arkansas at Fayetteville offers $40,000 over four years to students from Arkansas and Oklahoma. In 2003, the school announced ambitions to increase National Merit enrollment from 106 to 250 by 2010. That means recruiting and retaining about 63 a year. Already the school has risen from 92nd place in 2001 to 47th, with 47 freshmen this year. "We're competing with OU for these high-ability students in Oklahoma," said Suzanne McCray, associate dean of the Honors College. "We've moved up in U.S. News & World Report."
Given the incentives, separating colleges' interest and the public good isn't easy, noted Michael McPherson, a financial aid expert and president of the Spencer Foundation. "From a broad social point of view, I think it would be better if schools de-emphasized merit scholarships in favor of need-based scholarships," he said. But, as a former president of Macalester College, he understands why many don't. "There's a significant advantage in going out and buying National Merit Scholars rather than just buying kids with high grades or high SAT scores. You can take out an ad in the newspaper. If you're trying to brand the school as a place that has a lot of highly talented students, buying National Merit Scholars is a good way to do it," he said.