It was already evening when I suddenly remembered. I rummaged through my desk and then emptied a kitchen drawer, turning up several ambiguous candidates. Is a Woodclinched Eberhard-Faber Blackwing the same as a No. 2? My wife buys the pencils in our household, and she favors exotic leads. Suddenly gripped by an old but familiar fear, I hurried to the corner drugstore and bought a package of Harvard Squares.
College-bound high school students may have guessed the reason for my moment of panic: I had signed up to take the Scholastic Aptitude Test (SAT) the following morning, and a No. 2 is the only pencil whose mark is legible to the machines that score it. The instruction sheet I had received with my admission ticket advised me to bring two No. 2’s to the test. I sharpened four and put them beside my wallet on the bureau.
The SAT–published by the Educational Testing Service (ETS) under contract with the College Entrance Examination Board–is the most important college admissions test in the United States. More than 1.2 million high school students took it in 1983. I decided to take it again, a decade after I had taken it the first time, in order to learn first-hand how this test, which can determine the educational futures of the students who take it, is administered.
Early the next morning I left home without breakfast and joined two or three hundred teenagers on the sidewalk in front of Julia Richman High School on Manhattan’s upper east side. The school is an ironic symbol of American social progress: thirty years after Brown v. Board of Education, Richman’s students commute daily from Harlem to attend a black high school in a white neighborhood. But on days when the College Boards are given, white faces outnumber black ones. The sons and daughters of the privileged, who now attend private schools with names like Dalton, Trinity, and Collegiate, temporarily reclaim the building they have otherwise abandoned.
It was a frigid Saturday in December, but we weren’t allowed to wait inside the school. Shadowy figures occasionally peered out through the grated windows. About a dozen of the sort of students who volunteer to clean erasers after class were clustered around the locked front door, perhaps hoping to secure some small advantage on the test by thawing out first. Some students thumbed through test review books. I patted my pencils to make sure they were still in my pocket. And then–fifteen minutes after the announced starting time–the door creaked open and we filed inside.
“You will not be admitted to the test center without positive ID,” say the instructions ETS and the College Board give to students when they sign up to take the SAT. Test scores would be meaningless if college admissions officers couldn’t be certain they had been obtained under secure conditions. Students who lack driver’s licenses, passports, school identification cards, or other official documents with photographs attached to them are told to provide “a brief physical description of yourself on school stationary. It must be signed by you in the presence of your school principal or guidance counselor, who must also sign it.” Social Security cards, birth certificates, and parents’ driver’s licenses “are not acceptable.” In order to make certain all these instructions sink in, ETS repeats them when it sends students their test center admission tickets.
But on the morning I took the SAT, no one asked me for my identification. Indeed, when I took out my wallet to get my driver’s license, the proctor told me to put it away. She told all the students to put their identification away. “I just need to see your tickets,” she said. Proctors aren’t required to check the identification of students they know personally, but this woman couldn’t have known more than a few of the students in the room, because they were from different schools. She certainly didn’t know me. Nor did she seem to notice that I was ten years older than any of the other test-takers. I certainly would have asked for my ID.
Our proctor, who was wearing a jaunty scarf made of blue plastic netting, talked to herself as she waited for students to arrive. She told us she was going to give us our test booklets ahead of time but asked that we not open them “in case Dennis walks in.” Of course, several students opened their booklets immediately. Someone asked if we were allowed to use the booklets as scratch paper. The proctor said she didn’t know. It seemed to her that people had been allowed to last year, but that no one had said anything about it this year. We could probably “use the back,” she said.
This is incorrect. Students can use any part of the booklet as scratch paper. The are supposed to use the booklets as scratch paper. Our proctor would have realized this if she had looked at her instruction book, but she never did. Proctors are required to read several pages of fairly complicated instructions aloud to students, asking them periodically whether they understand. But our proctor didn’t read a word.
Shortly after she gave us our booklets, she told us to begin. We would have thirty minutes to complete the first section, she said, starting now. Then, after we had started, she told us to be certain to fill in the identifying information on both sides of the answer sheet and the back of the test booklet. Students have to provide quite a lot of information: name, address, birth date, two signatures, sex, test form number, registration number, form code, test book serial number, test site location, test site code. Students fill in much of this information by writing it in little squares and then darkening boxes on the answer sheet that correspond to individual letters and numbers. Doing all this takes a long time. Students are supposed to be allowed to do it before the test begins, with the proctor leading them through every step of the way and checking to make certain the test and booklet numbers are entered correctly. The students in my room were cheated out of perhaps one-third of the time allowed for the first section of the test. Even students who ignored the proctor and began working were penalized, because the proctor talked continually. One thing she talked about was the serial number: she wasn’t sure if there was one, or, if there was, where in the world it might be.
There was no clock in our room, so the proctor periodically marked the time on the blackboard. Her timing was very approximate. I checked her with my watch. She shaved off a few minutes on some sections, added a few minutes on others. My desk was so covered with graffiti that the ink from the tabletop sometimes rubbed off on my answer sheet. Erasing these marks was difficult. All the desktops had been carved and gouged. It was possible to tear an answer sheet simply by marking answers.
Proctors are explicitly required to give students a five-to-ten minute break at the end of each hour of testing time. Our proctor gave us only one break, very late in the test, and she gave it to us only because a student complained. Several students continued working on their test during the break. This, of course, is against the rules. It is cheating. The proctor said nothing, even though she was clearly able to see what was going on. Other students worked on sections other than the one they were supposed to be working on. This, too, is cheating.
Students who finished the last section early were allowed to leave before the end of the test. This is absolutely forbidden. It’s, also extremely distracting to the students still working. The students who left early rustled their coats and papers and talked in normal voices. The proctor talked, too. She stood in the doorway and talked to a woman in the hall. I was working on the last few problems in my second math section at the time. Every time the proctor or one of the students started talking, I lost track of what I was doing and had to begin again. When time was finally called, the proctor allowed the remaining students to continue working on the test. Several were still working when I left.
I was appalled by the way the test had been run. In January I wrote a letter to ETS describing what I had observed at Julia Richman High School. I assumed that ETS would cancel the scores of all the students in the room, if not in the entire test center, and schedule a retest. At the very least, I thought, ETS would contact the other students in the room to let them know they had been given the test improperly and to offer them an opportunity to take it again.
For nearly two months, I heard nothing. Then, finally, I received a letter from Patricia A. Goccia in the “Candidate Relations” office of ETS’s Admission Testing Program. Goccia said she was sorry I had found the testing conditions “unfavorable.” She said, however, that there was “no indication on the Supervisor’s Irregularity Report informing us of the situations described in your letter.” She said she had called the supervisor just to make sure. (Every test center has one supervisor, who is responsible for keeping track of the proctors.) This supervisor, she said, “confirmed” that the identification of all students had been checked and that all other rules had been followed.
“In concluding my investigation,” Goccia wrote, “the supervisor assured me that all procedures were followed as outlined in the Supervisor’s Manual, and feels that the instructions were in no way breached.” As for the cheating I had seen, and the failure of the proctor to read the instructions, and the misinformation given to students, and the failure of the proctor to time the test properly, and the many flagrant violations of official testing procedure that I had described–Goccia had no comment. She simply did not acknowledge the substance of my letter. “Had you felt that you did not test to your utmost ability under optimum conditions,” she reminded me, “you did have the option to cancel your scores by Wednesday following the test date.”
I called Goccia the next day. She was very nice. She sounded young. She was trying to be helpful.
“The supervisor checked with the proctor,” she said, “and to her recollection, she said the. administration went perfect.” Goccia said that my proctor had been “doing the test for six to eight years, so the supervisor felt that she was competent enough. She had never brought any concerns to his attention as far as instructions or pre-admission procedures.” She said that the supervisor himself had been giving the tests for half a dozen years, and that “he has never had complaints such as this.”
All this apparently reassured ETS. But it made me wonder what had been going on at Julia Richman for the last “six to eight years.” The infractions I had observed were apparently standard operating procedure.
I asked Goccia if it was a violation of the rules to make students fill in their names and other information when they were supposed to be taking the test. “It’s not really a violation,” she said, “as long as it’s not included in the testing time.” But it had been included in the testing time, I said. “Then it would be a violation.”
I asked her if students were allowed to work during breaks. She said no. I asked her if students were allowed to continue working after the end of the testing period. She said no. I asked her if students were allowed to work on any section they wanted to. She said no. I asked her if students were allowed to leave before the end of the test. She said no, that was definitely a violation of the rules. She also said that reading instructions aloud and checking identification were both “mandatory.” Failure to do either would be a violation of the rules. Hourly breaks are also mandatory, according to the rules, although Goccia said this requirement wasn’t strictly adhered to. “I remember one case,” she said, “where the supervisor just said he will not give breaks anymore, because he had windows broken and supervisors beat up, and he just didn’t want to have kids go out of the classroom.”
For all of this, Goccia seemed bizarrely uninterested in what I told her. Indeed, she was confused as to why I had called. She reminded me once again that if I really thought the testing conditions were inadequate, I should have canceled my scores. I told her I didn’t want to cancel my scores; I wanted her to cancel everybody else’s scores. After all, every student in the room had been cheating, or being cheated, or both.
ETS knows that cheating is a problem on its tests. “Many types of cheating are currently not detected,” said an internal ETS “briefing paper” in 1980. But the document also noted that the cost of detecting cheating had been rising in recent years (from $136,000 in 1978 to roughly $300,000 in 1980) and that “[t]his substantial increase in expense calls into question again the cost/benefit ratio of the test security function.” One possible solution would be “[d]iscontinuing the test security function altogether”–an attractive option because “[test security functions at ETS result in substantial amounts of negative publicity.”
Consistent with this concern for its publicity, ETS had referred my letter not to its Test Security Office, the department charged with detecting cheating, but to Candidate Relations. Goccia’s job was to calm me down, not to correct an injustice. I asked her what ETS was supposed to do in cases where it “knows that procedures have been violated and that cheating has taken place.”
“Well, see,” she said, “we have a specific department that deals directly with the supervisors. My job is just basically to handle our candidate inquiries. After I did my investigation, I passed it on to our Supervisor Relations Department, who will take any further action if necessary. I mean, it’s going to take the supervisor coming out and saying, ‘Yes, this did happen, we didn’t do this right’. And he’s not doing that.”
When I persisted, Goccia said that she would check with Supervisor Relations and see if anything should be done. She called me back later to say that ETS had consented to send an observer to Julia Richman the next time a test was administered there, which would be in May of 1984. Since this was clearly being done only because I had continued to press my case, I decided to be there, too.
“We hear war stories from the kids all the time,” says John Katzman, a young New Yorker who runs a highly successful SAT coaching school called the Princeton Review. “It is unbelievable how bad some of ETS’s test centers are. There was a proctor at Stuyvesant who didn’t even bring a watch. So, at some point in the middle of the first section of the test, he asked the students, Does anybody know what time it is? And one of the kids said, Nobody has a watch; why don’t you go look at the clock down the hall? And he did. Every time he went to see what time it was, he left the kids all alone in the room and they could do whatever they wanted.”
I asked Katzman why there was such a problem.
“What you have to understand,” he said, “is that these proctors get paid $30. When you figure out all the time they’re supposed to spend, that works out to just about the minimum wage. And yet if they lose a test booklet, ETS is after them for the rest of their lives. If they turn in someone for cheating, they have to do a lot of paperwork, and then the kid threatens to kill them, all for thirty bucks. The proctors just don’t take it seriously. They’re in a coma.”
No Bathroom Supervision
Some proctors turn the other way while students consult pocket dictionaries or work math problems on wristwatch calculators. Students too timid to cheat in the testing room wait for a break and then use their dictionaries and calculators in the bathroom. (ETS’s security budget doesn’t provide for bathroom supervision). I’ve heard stories of testing centers where students flipped through the pages of their test booklets before the test, calling out words they didn’t know while students with dictionaries looked up definitions. I’ve heard about centers where students actually walked around the room during the test, comparing their answer sheets with those of other students. I’ve talked to six people who told me they’d been paid to take the SAT for other students; four of these impostors were paid directly by the parents of the students for whom they had been cheating. I heard about one girl who, after doing very well on the PSAT, became the object of a bidding war among classmates willing to pay for the privilege of sitting next to her during the SAT.
One recent fad in cheating on the SAT is taking the test untimed. ETS permits students with “diagnosed visual, physical or perceptual handicaps of a temporary or permanent nature” to take the test over periods as long as two days. Many students legitimately need this service. But others see it as a golden opportunity to improve their scores, and there are doctors willing to say that almost anyone has a “perceptual handicap.” I heard about two friends (one with a legitimate handicap) who took the SAT untimed about ten days apart. ETS used the same test form both times. The first student to take the test consulted a dictionary during breaks. The second learned many of the questions in advance from his friend.
I took the SAT again at Julia Richman in May. Once again, no one checked my identification. My proctor this time was a young woman who was having a great deal of trouble reading her seating chart. She couldn’t decide which end went where and she couldn’t figure out how the diagram related to the classroom. She turned it upside down, then right-side up, then upside down. I simply found my name on the paper and took my seat. She didn’t ask me who I was. I didn’t even take my wallet out of my pocket.
The proctor was having so much trouble with the seating chart that finally the supervisor came in and told her just to seat us anywhere. A few students changed seats. The supervisor turned on the lights, something our proctor had forgotten to do, and left.
Then Anella Schmidt, the ETS observer, arrived. Schmidt was a solidly-built woman with the stern look and patronizing demeanor of a minor government official on a tour of the provinces. Her gray hair was tied back in a bun. She wore a name tag identifying her as an emissary from “ETS, Princeton, NJ.” Periodically she made portentous jottings in a green notebook. Her first act was to modify the supervisor’s random-seating order. She explained that ETS needed to know where everyone was sitting so that it could later detect cheating. “Are any of these Perezes related to each other?” she asked, scowling at the seating chart. Then she went around the room noting where everyone was sitting.
Our proctor wasn’t quite sure what to do. This was her first time giving the test. She apparently hadn’t read her instruction manual, although she had certainly done something with it: It was torn, folded, and crumpled. Schmidt looked over her shoulder and told her what to do. “You’ve got to read all that,” she said when the proctor skipped part of the instructions. By this time Schmidt had taken the seat next to mine. I may be paranoid, but I find it peculiar that, with several hundred students and more than a dozen classrooms to choose from, Anella Schmidt decided to sit next to the person whose phone call had prompted her trip. If I were Anella Schmidt, I’d have wanted to observe the proctor about whom the complaint had been made, not the person who had made it.
It was hard to hear the instructions, because the proctor had a quiet reading voice. She stumbled over some of the longer words. When she got to the part about using test books for scratch paper, she inserted the word “not” and made it sound as though we weren’t supposed to. There was some confusion about this; eventually it was cleared up. Then a girl in the back of the classroom asked if we were supposed to guess. The proctor deferred to Schmidt.
“I wouldn’t guess unless you can eliminate two of the five,” said ETS’s official representative. “Guessing isn’t going to help you.”
As ETS knows, this is absolutely untrue. Because of the way the SAT is scored, students who can eliminate even one choice on a problem are most likely to improve their scores by guessing blindly among the remaining answers. Guessing intelligently is one of the keys to doing well on the SAT. Anella Schmidt of “Princeton,” on hand to ensure that the test was properly administered, misinformed us about how to take the test.
Schmidt left when we began to work. Our proctor spent most of her time standing at the window, with her back to us, watching some boys play basketball (noisily) on the playground below. It was a nice day and she didn’t want to be doing what she was doing. At break time students were excused, in groups of two or three, to go to the bathroom, which, as usual, was unproctored.
I talked to Katzman a couple of days after the test. He said that a bored proctor at one of the test centers hadn’t timed students on individual sections of the test, but had merely told them they had three hours to complete the whole thing. This meant that the students could work on any section they wanted to whenever they wanted to–an enormous advantage. Another proctor, for an Achievement Test, had fallen asleep at his desk. One hour went by. Then another. When he finally woke up, his students had been happily working on a one-hour exam for nearly three hours. The proctor asked the students not to say anything, because he didn’t want to get in trouble with ETS.
“The SAT is an unproctored exam,” Katzman says. “You take it on the honor system.”
©1984 David Owen
David Owen, formerly a senior writer for Harper’s Magazine, is investigating standardized testing and American education.