Willis W. Clark was a quiet, scholarly man who liked almost nothing better than tinkering with his theories about measuring the abilities and achievement of students. His wife, Ethel M. Clark, on the other hand, found the subject of testing rather boring. But when educators outside Los Angeles started expressing interest in one of the tests created by her husband, the assistant director of research for the Los Angeles public schools, Mrs. Clark's imagination soared.
The district wasn't in the business of marketing tests, she knew, so why not set up a commercial concern and become the middleman, or the middlewoman?
Mrs. Clark worked up the courage in 1926 to approach her husband's boss, the district superintendent, to ask for the rights to the Los Angeles Diagnostic Tests in the Fundamentals of Arithmetic. Permission granted, she worked out a royalty arrangement with the district and then proceeded to do market research that has become legendary. Mrs. Clark sent out penny postcards--that's all they cost back then--to 25 large districts, promoting the availability of the test.
And then she waited. And waited.
Nearly a year went by, according to the Clarks' daughter, June Duran. "She had practically forgotten about them," Duran recalls.
Ethel Clark would probably have moved on to her next money-making scheme, of which she had several. But then an order came in.
The Kansas City, Mo., district wanted 20,000 copies of the test. The California Test Bureau was in business. From these humble beginnings, the Clarks helped shift the field of student testing from the province of academe into the commercial marketplace.
To educators around the country, the little company that was capitalized, in a sense, on a 25-cent investment in postcards, is now known as CTB/McGraw-Hill, the testing division of a publishing giant, the McGraw-Hill Companies Inc.
Known for its California Achievement Test and the Comprehensive Test of Basic Skills, CTB/McGraw-Hill is one of the big three purveyors of K-12 tests, along with Harcourt Educational Measurement, the San Antonio-based publisher of the Stanford Achievement Test and the Metropolitan Achievement Tests, and the Riverside Publishing Co. of Itasca, Ill., which publishes the Iowa Tests of Basic Skills. These companies have prospered for decades by offering off-the-shelf standardized tests.
"It is a relatively small segment of educational publishing," George F. Madaus, the director of the Center for the Study of Testing, Evaluation, and Educational Policy at Boston College, says of the market for precollegiate tests. "But it's lucrative. It must be, or else they wouldn't be in business."
But the landscape has changed in the past decade, as more states have adopted high-stakes accountability programs with their own academic standards and assessments. To be sure, the major school test publishers have stepped forward to bid on developing and processing these customized tests. But such specially crafted state tests are more expensive than the off-the-shelf variety.
Commercial publication of tests began before the First World War. In 1916, Houghton Mifflin Co., a leading textbook publisher at the time, began publishing the Stanford-Binet intelligence scale, an individual-intelligence test created by the psychologist Lewis M. Terman of Stanford University. The test quickly took hold in schools.
In addition to intelligence tests, the schools had dozens of achievement tests and "scales" written by university professors or school district research directors. Many of them were published for sale around the country.
At first, sales were handled by the test developers themselves, or by university research bureaus. In 1918, the yearbook of the National Society for the Study of Education included a report about educational tests.
It listed dozens with such distinctive names as Courtis' Standard Research Tests in Arithmetic, Ayres' Handwriting Scale, Woody's Arithmetic Scales, Murdoch's Scale for Measuring Certain Elements in Hand Sewing, and Brown's Connected-Latin Test.
The education society noted in its yearbook that "in only a few cases are the tests published on a commercial basis, and for this reason, one may be reasonably certain that the tests may be obtained at approximately the cost of printing."
In 1922, a writer in The American School Board Journal observed: "Within the past five years, so many tests have been produced that the teacher who wishes to secure a measure of her instruction frequently is at a loss to know which of a great number of tests to select. For general circulation, there are now 31 tests in reading, 30 in language, 26 in arithmetic, 24 in handwriting, and 17 in spelling."
Although Houghton Mifflin published the Stanford-Binet individual-intelligence test--and still does--the company passed up the chance to publish the first group-intelligence tests, which, unlike the Stanford-Binet, could be administered to many students at once.
When the company declined, Terman and his associates turned to another textbook publisher with which he had dealings, the World Book Co. of Yonkers, N.Y.
World Book was founded in 1905 by Caspar W. Hodgson, a Stanford University graduate who became a school principal in Indiana and California. A former textbook salesman for D.C. Heath and Co., he started his own publishing company at the urging of the schools superintendent for the Philippines, who had complained about the unsuitability of the textbooks available for his system.
The World Book Co., which was not affiliated with the famous encyclopedia, built its niche with elementary school texts--and with tests.
"Hodgson really saw the potential of group tests," says Henry L. Minton, a professor of psychology at the University of Windsor in Ontario and a biographer of Terman.
Before World War I, World Book published an intelligence scale designed by Arthur S. Otis, a student of Terman's at Stanford. After the war, the company published the National Intelligence Tests, a creation of the same group of psychologists that developed the famed Army Alpha tests.
World Book also published Terman's own group-intelligence tests, which the Stanford professor had formulated in 1919 for grades 7-12.
Terman had become a sort of scholarly adviser to World Book by the early '20s. Hodgson even asked him to take a leave of absence from Stanford to become the company's test expert.
"If we could have such a man as you for about one year to get us headed right as an expert in tests and measurements and to some extent as a psychological editor of textbooks in general, we would feel that we might get a younger man started with you before you get through to carry on the work," Hodgson wrote Terman in June 1921. The letter is one of many in Terman's papers archived at Stanford that help shed light on the commercial origins of testing.
Terman declined the offer, telling Hodgson he was too busy and couldn't possibly take a leave. But he did agree to become the general editor of the Measurement and Adjustment Series, a set of books and monographs that was meant to advance the testing movement among classroom teachers and school administrators.
In 1921, Terman wrote to World Book to propose "an all-round achievement test covering in three- to eight-page test booklets all the school subjects from grade three to grade eight."
"The average superintendent is rather lost among the great numbers of tests for every sort of purpose," Terman added. "It costs a good deal of money and more time to test out all the different subjects of the curriculum at the present time because there is no one satisfactory battery of tests covering the whole ground. I am convinced the time has come for such a battery of tests."
Thus began the Stanford Achievement Test, one of the most frequently administered school tests today.
Of course, Terman explained, such a battery would be more complicated to write than his previous tests. He and his co-authors at Stanford demanded more than the standard 5 percent royalty on tests. They ended up getting 9 percent, just shy of the 10 percent World Book paid textbook writers.
As Terman and his colleagues were refining the tests, he and the publisher began to bandy names around. Hodgson suggested the Terman Achievement Test.
"The only objection to using the name Stanford is that it might prejudice its sale a little in California among the University of California people who predominate in the schools," a market-conscious Hodgson wrote to Terman.
Terman replied, without elaboration, that he preferred the Stanford name.
Annual sales of the Stanford Achievement Test exploded from 115,000 in 1923, its first year, to 1.5 million by 1925.
That a textbook company would also publish and sell intelligence and achievement tests might be called synergistic today. But in the field's infancy, it appears that educators paused to consider whether such an arrangement was appropriate.
According to a letter in Terman's papers, the brand-new American Educational Research Association debated the commercial publication of tests sometime in 1918 or 1919. In a 1952 exchange of letters, a longtime World Book employee, O.S. Reimbold, tried to get Terman to help him remember the details.
"[Arthur] Otis and other authors who had contracted with World Book Company for the publication of their tests were given a vote of censure," Reimbold recalled in his letter. But a year later, after World Book had published its first tests, "the same AERA voted not only approval, but an expression of satisfaction with the product," Reimbold wrote.
Besides displaying the tests at the major education conventions of the day, World Book and other publishers used their network of textbook salesmen to push tests throughout the market.
They also employed advertisements. A 1923 ad in the School Board Journal touted the Otis Classification Test. A classroom set of 25 exam booklets and the scoring key cost $1.30. A manual was available for an additional 25 cents.
By that time, Willis Clark was hard at work in Los Angeles on one of those many single-subject tests that Terman was trying to replace.
When the order came in from Kansas City for 20,000 copies of the arithmetic test, Ethel Clark went on a frantic mission to find a printer to fill it. "The first thing they asked me was how I was going to pay for it," she says in a company history compiled by her daughter. "Well, that was something I hadn't thought about, so a second mortgage on our house paid for the 20,000 tests."
Mrs. Clark and her mother, Jeanette King, stored the printed tests in the family garage. When the tests were ready to ship, the two put them in a big piano box. The package was so heavy, the bottom fell out. Repacked in smaller boxes, the tests eventually arrived in Kansas City.
With training from a secretarial school, Mrs. Clark had long helped supplement her husband's income by typing master's theses and doctoral dissertations. Because of that experience and the expense of printing out-of-house, Mrs. Clark decided her fledgling company could print the tests itself. So she bought a printing press, a purchase that came in handy when the company found itself doing more business printing business cards and stationery than tests, says Duran, the Clarks' daughter.
Early in the life of the California Test Bureau, Mrs. Clark came up with another idea for making money from schools. State textbooks were distributed from a depository in San Francisco, even though the bulk of student enrollment was in Southern California districts, which paid an extra 5 percent in shipping fees. Upon learning this, Mrs. Clark opened the Southern California Book Depository in 1930.
"She was off on another venture," her daughter says. At times, the testing business became "sort of a stepchild," Duran says.
Willis Clark continued working for the Los Angeles schools, and he continued writing tests. In 1933, he and Ernest W. Tiegs of the University of Southern California developed a test battery known as the Progressive Achievement Tests.
"The name really wasn't tied in to the progressive movement" in education, Duran says. The tests were well-received in schools, and Ethel Clark retrained her focus on test publishing.
About the same time, another invention helped advance the testing business. With the advent of the multiple-choice test around 1915, students circled their answers on a test form or booklet. Scoring was tedious, even with stencils that were laid over the answer sheets to speed up the process.
Around 1936, the International Business Machines Corp. created an electronic test-scoring machine that made grading multiple-choice tests more efficient.
Duran remembers the machine because, as a youngster who worked after school in her mother's business, "I was the scoring department," she says.
"The machine was as big as a desk, and the procedure was to take an answer sheet and drop it in the slot, one at a time," Duran says. "The graphite [from pencil marks] that came through produced a little electrical charge, which was indicated on a volt meter on the desk. You had to drop every one of those damn tests in the slot, but it was an improvement."
As Duran recalls, her parents were not getting rich from the business. When schools didn't pay their bills right away, the couple took second and third mortgages or used the accounts payable as collateral for loans.
World War II proved especially challenging. The U.S. Army decided it needed CTB's Los Angeles facility and ordered the company to vacate the premises within 30 days. The company had a huge book inventory, as well as printing presses, linotype machines, and other equipment.
Mrs. Clark sold off the book-depository inventory, and the family was left with its test business and a staff depleted by the war. CTB moved to smaller quarters on Hollywood Boulevard.
Other challenges had also emerged for test publishers.
In the late 1930s, a Rutgers University professor, Oscar K. Buros, began publishing the Mental Measurements Yearbook, a sort of Consumer Reports for the users of standardized tests.
He asked reviewers to examine the popular tests of the 1930s and published their comments in the 1938 yearbook, causing a stir in publishing circles. Time magazine reported that test publishers appealed to Buros "in the name of common decency" to stop the presses. "A distraught publisher: 'Now, Oscar! Is this sporting?' "
The Mental Measurements Yearbook is still published by the Buros Institute of Mental Measurements at the University of Nebraska at Lincoln.
In a lecture shortly before his death in 1978, Buros lamented what he viewed as a lack of substantial progress in educational tests since the 1930s. "Today's tests are more attractively printed and are generally machine scorable, but otherwise, they show relatively little improvement," he said.
Buros did, however, single out for praise work at the University of Iowa.The Iowa testing program started out in 1928 as a state academic contest open to all high school students. An education professor at the university, E.F. Lindquist, became the director of the program, and the tests became known as the Iowa Every-Pupil Tests. Later, Lindquist and his colleagues created the Iowa Tests of Basic Skills and the Iowa Tests of Educational Development.
The Iowa tests were gaining popularity in neighboring states, so in 1940, Lindquist reached an agreement with Houghton Mifflin to distribute them. The company that published the most popular individual-intelligence test but had spurned group ones in the early part of the century was now a player in the growing achievement-test market. The University of Iowa was, and remains, the recipient of royalties for the Iowa tests.
Lindquist also assisted with test-scoring. Although IBM had come out with an optical-scoring reader in 1936, which speeded up the process, it was a relatively slow machine. Lindquist liked to tinker at night, and he devised a plan for an optical-scoring machine that would read specially printed score sheets at a rate of hundreds per hour.
Over at the World Book Co., times got a bit tough during the Great Depression, especially for book sales. It didn't help that company President Caspar Hodgson was diverting profits to his own use, which included paying off a blackmail threat involving his mistress, according to Minton, the biographer of Lewis Terman.
Terman and other test authors were concerned about their royalty payments, but the publisher eventually got its house back in order. In 1934, World Book sent Terman an advertising circular extolling the fact that 24 million Stanford Achievement Tests had been sold. World Book considered the test "one of our mainstays, and [we] believe it will be profitable for authors and publishers for years to come," it wrote Terman.
Later that year, Terman wrote to World Book and evidently expressed some concern about a new test the company had begun publishing, the Metropolitan Achievement Tests. The MAT was written for use in the New York City schools, and World Book was making it available as a competitive offering even with its own Stanford tests.
A World Book executive tried to allay Terman's concerns. "In view of our experience, we felt that we should have a new achievement test to satisfy that class of school people who are always on the lookout for something new," the executive wrote. Had World Book not published the Metropolitan, "it is very likely that either the New York City board or some other publisher would have done it," he added.
The Stanford Achievement Tests "seem to be as much in demand as ever," the letter said.
As things would turn out, the Stanford and Metropolitan achievement tests would outlast the World Book Co. By 1960, the clubby world of book publishing was undergoing a transformation from small, privately held firms into public ownership and mergers.
World Book could no longer survive on its own. In 1960, it merged with the venerable publisher Harcourt, Brace & Co. to form Harcourt, Brace & World Inc. The merger was a good fit. Harcourt was a leader in high school- and college-textbook publishing, while World Book still had many elementary texts along with the Stanford, the Metropolitan, and other tests.
By the 1950s and '60s, change was buffeting the California Test Bureau as well. Although the company had survived World War II and the loss of its book-depository business, it faced a new threat in the early 1950s. June Duran says that her mother always paid union wages to her print and bindery workers, even though she insisted on maintaining an open shop.
Nonetheless, the International Typographers Union attempted to organize workers in 1951 and demanded that Ethel Clark sign a union contract.
"There was no way in hell my mother was going to let anyone organize the company," Duran says.
Mrs. Clark and others at CTB packed up tests and other inventory and shipped them off to branch offices in Madison, Wis.; Dallas; and New Cumberland, Pa. The union began 11 months of picketing CTB's Los Angeles headquarters, but Mrs. Clark organized caravans to bring her workers across the lines.
The union, says Duran, alerted fire marshals about the piles of paper scraps that were typical of any printing operation. So Mrs. Clark had her workers pack up the paper trimmings and take them home at night.
The labor battle subsided, but CTB was changed, Duran says. Printing was moved permanently to the Midwest, with more shipping and scoring services performed by the branch offices as well.
By the late '50s, CTB was outgrowing its offices on Hollywood Boulevard. The Clarks were considering other properties in the area when it came time for their annual retreat to their summer home on the Monterey peninsula. CTB had held its annual sales conferences at the Del Monte Lodge in Monterey. Why not move the whole operation to such a beautiful area, the Clarks thought.
After finding a suitable site in a new research park, Ethel Clark returned to Los Angeles to tell her workers about the move. She gave them two weeks' paid vacation to check out Monterey.
"She couldn't imagine anyone not wanting to move there," Duran says.
In September 1960, trucks carried some 300 tons of tests, files, desks, and 24 IBM scoring machines to a new, $375,000 headquarters overlooking Monterey Bay. But of some 60 to 70 CTB employees who were expected to move, fewer than 30 actually did.
CTB struggled through scoring season that fall. Finding employees was harder on the less populous peninsula. Duran recalls that many military wives from nearby bases were hired temporarily.
The company faced other hurdles. Early in the 1960s, it received a line of credit from a local bank so work could proceed on the updated standardization of its tests. But the bank changed hands, and the line of credit was withdrawn. Finances were again tight.
Willis Clark's health began to deteriorate, and some employees became apprehensive about the future of the company, Duran says. They tried to buy out the Clarks, but couldn't raise enough money. Longtime loyalties within the close-knit company were being split.
John Stewart, who joined CTB in 1963 as an "evaluation consultant," the firm's term for sales representative, remembers that the company was struggling. "It was on pretty shaky ground," says Stewart, who had worked for the Illinois education department before taking over a huge Midwestern sales territory for CTB.
Willis Clark died of a cerebral hemorrhage in 1964 at the age of 69. Ethel Clark, then 66, worried that she would not be able to continue running the company. Though CTB had received inquiries previously, Mrs. Clark had not been interested in selling. In her behalf, friends began making overtures to purchasers, including McGraw-Hill.
In June 1965, McGraw-Hill purchased Ethel Clark's 81 percent interest in CTB for $200 a share. The overall selling price wasn't disclosed. With roots in business and technical books and magazines, such as Business Week, McGraw-Hill was a newcomer to educational publishing in the 1960s.
When McGraw-Hill took control of CTB, Duran was offered a management position in the new division. "I couldn't believe they wanted me," she says. "It was the highlight of my life."
She worked as the managing editor of tests, then oversaw intellectual-property issues and author contracts. She became active in the test committee of the Association of American Publishers.
Ethel Clark, for reasons June Duran does not fully understand, did not approve of her daughter's staying on. Mrs. Clark died of cancer in 1969.
McGraw-Hill's purchase of the California Test Bureau came at a propitious time. In 1965, Congress enacted the Elementary and Secondary Education Act, which included Title I. The compensatory education program funneled federal dollars into education on a grand scale and called for "program evaluation." That meant a significant increase in norm-referenced tests--those that compare student against student.
"I remember that when Title I went into effect in 1966, schools called up CTB and bought the catalog," says Ross Green, now a senior research manager there.
"The accountability push led to a lot of developments," he adds.
For one, CTB rolled out its first major new test in many years soon after being acquired by McGraw-Hill. The Comprehensive Test of Basic Skills became as well-known and as widely used as the company's longtime flagship, the California Achievement Test. Today, the CTBS is the core assessment in the company's 2-year-old TerraNova program.
By 1968, CTB/McGraw-Hill, as it was called by then, had annual revenues of $2.8 million.
Other mergers were occurring. In 1970, Harcourt purchased the Psychological Corp., which had been formed in 1921 by eminent psychologists, such as Edward L. Thorndike of Columbia University and Lewis Terman, to make tests for practicing psychologists. By the time Harcourt bought it, Psychological Corp. had annual revenues of $5 million.
Harcourt, of course, already had a testing division from its merger with World Book, and it already owned major school tests, such as the Stanford and Metropolitan achievement tests. Psychological Corp. added a whole catalog of tests such as the Wechsler Intelligence Scale for Children, probably the most widely used intelligence test after the Stanford-Binet.
Today, Harcourt Educational Measurement, based in San Antonio, is the educational testing division of Psychological Corp., which in turn is part of the publishing and media conglomerate Harcourt General Inc.
Meanwhile, Houghton Mifflin renamed its test subsidiary Riverside Publishing Co. in 1979. The company wanted a separate name for its testing business, "to avoid the perception in the marketplace that a school needed to use Houghton Mifflin instructional materials in order to use Riverside Publishing's tests," according to a company history.
The distinction is one that McGraw-Hill and Harcourt have never copied, given the close links between those companies' testing divisions and their parent companies. At times, they have faced questions about the perceived alignment of the tests and their parents' curriculum materials.
For example, Harcourt came under fire in Texas this year because of a brochure advertising its math texts and other educational materials as "the only program to have tests written by the same company that helps to write the [Texas Assessment of Academic Skills] test." Harcourt is a subcontractor for the state's basic-skills test.
A decade ago, similar questions were raised about a test-practice kit sold by McGraw-Hill, which contained the same or similar questions as those on CTB's California Achievement Test.
That and other criticisms of the commercial test publishers have done little over the years to lessen schools' reliance on the leading standardized tests. In fact, as the accountability movement took hold in the late 1980s and 1990s, the role of the three major test publishers grew, even though they have had to retool their products to meet states' changing demands.
A number of states had long prescribed the use of the publishers' off-the-shelf tests as their de facto state achievement tests. CTB had a state-contracts department as long as two decades ago. Back then, creating a state achievement test was no more difficult than "taking a shelf product and putting a custom cover on it," says Linn Williams, a CTB national accounts manager.
But as states wrote their own academic standards, many wanted their tests to better reflect those expectations. They wanted customized tests.
Some states tried producing their own. But the costs, as well as concerns about reliability and validity, dampened their enthusiasm. Meanwhile, the commercial publishers responded quickly and began marketing their ability to produce customized assessments.
Today, CTB/McGraw-Hill is the publisher of state assessments in 22 states, including New York, Colorado, and Indiana.
The competitiveness of the market these days can be seen by examining CTB's native California.
In 1997, the state solicited bids for a standardized-testing program after its pioneering performance-based version was felled by political conflicts. CTB, Harcourt, and Riverside Publishing bid on the contract.
The state had authorized spending as much as $35 million for the first year of the program, or $8 per student. Riverside bid $8; CTB, $4.85; and Harcourt, $2.89.
After getting the advice of dozen of educators, state Superintendent of Public Instruction Delaine Eastin recommended CTB's Comprehensive Test of Basic Skills for a one-year contract, although she was unenthusiastic. All three submissions, she said at the time, "were seriously flawed, with the level of test content falling years below grade-level expectations."
In its proposal, CTB had emphasized its California roots, as well as the use of its tests in 70 percent of the state's districts. It had reserved 4 million pounds of paper and space at five printing plants for the booklets and answer sheets that would be used to test California's 4.3 million pupils.
When the state board of education met that November, CTB was in for disappointment. The board voted 9-2, to award a five-year contract to Harcourt Brace. Beginning in 1998, the state's students would all be required to take the Stanford Achievement Test, not the TerraNova.
In 1998, CTB bounced back from its defeat by winning a four-year, $30.5 million contract to produce Kentucky's tests. CTB is heading up a consortium that includes three smaller companies.
The Kentucky contract highlights the fact that CTB, Harcourt, and Riverside may dominate the off-the-shelf-testing business, but other players are elbowing their way into the growing market for customized state tests.
The most prominent is Minneapolis-based National Computer Systems Inc. A leading test-scoring company, it has won the testing contract in Arizona, with CTB as a subcontractor, to create the test.
"There are a number of smaller companies that will bid on these state contracts," says Madaus of Boston College. "I don't think they will get to the size of the big three, but they are real players."
Just how big the big three are today is not easy to measure, given the way in which their parent companies break down finances. But Simba Information Inc., a Stamford, Conn., business-research firm, estimates that CTB/McGraw-Hill, Harcourt, and Riverside Publishing will have combined sales this year of $221 million.
In a 1997 report, Simba estimated that CTB had K-12 test sales of $80 million in 1996, Harcourt had $50 million, and Riverside had $48.7 million, for a total of $178.7 million.
A CTB newsletter for employees says that in 1998, the company "achieved a revenue milestone--exceeding $100 million in sales."
"Right now, there is a great demand out there for assessment," David Taggart, the president of CTB/McGraw-Hill, says. "The challenge for remaining profitable is figuring out how to remain competitive as we go from shelf business to custom business. This business has become driven by the states."
CTB has come a long way from the 25 penny postcards that Ethel Clark mailed in 1926.A few years ago, it moved into new headquarters in Monterey, a huge building on a hilltop that overlooks the bay. The facility is home to its test-development and -research operations. It also handles a small portion of CTB's scoring and reporting tasks.
CTB sold 10 million tests last year and scored twice that many, says Michael H. Kean, the company's vice president for public and governmental affairs.
June Duran still lives in the Monterey area and sometimes finds herself visiting the headquarters of the company her mother founded when June was only 6.
She thinks back to when her father became interested in testing. As a sociologist at the Whittier State School for Boys, Willis Clark got hooked on the idea of writing "case histories" of each child in an effort to discover the cause of delinquency.
He became convinced that his theories about curtailing improper behavior could be applied to the vast numbers of students in the public schools through the use of diagnostic and achievement tests.
Today's high-tech score reports, with their detailed information about a child's strengths and weaknesses, remind Duran of her father's original goals.
She also thinks about her family's business struggles and how CTB probably would not have survived without the McGraw-Hill purchase.
"McGraw-Hill really pumped the money in, and the personnel, and everything it needed to be a success today," she says. "When I go by the building today, I just feel very proud."
|When educators outside Los Angeles expressed interest in one of the tests created by her husband, Mrs. Clark's imagination soared.|
Vol. 18, Issue 40, Pages 30-36