History of the GMAT
In March 1953, deans from Columbia, Harvard, Northwestern, Rutgers, Seton Hall, University of Chicago, University of Michigan, University of Pennsylvania, and Washington University (St. Louis) met with representatives from the Educational Testing Service (ETS) to create an objective, national entrance exam for graduate business school. The exam would test relevant skills that lead to success in the core curriculum of graduate business schools. It would be long enough to be statistically reliable and would be administered and scored uniformly.
In February 1954, ETS administered 'The Admission Test for Graduate Study in Business' (ATGSB) to 1300 students. In the first year, ETS administered the ATGSB to 4288 test takers. At that time, only eight business schools required the exam as part of the application for admissions. Today the GMAT exam is required by more than 1800 business schools and the number of tests administered in 2007 had grown to 219,077.
Changes in the Administration of the GMAT
In the fifty-five years since the first meeting of the Graduate Business Admission Council (changed to the Graduate Management Admission Council in 1976), nearly every aspect of the test has changed. In 1976, the name of the exam changed from the Admissions Test for Graduate Study in Business (ATGSB) to the Graduate Management Admission Test (GMAT). In 1955, one year after the exam was first administered, the Quantitative and Verbal sectional scores were added. In 1961 all but one of the question types were changed and they continued to change until 1994. The length of time required to take the test grew from two hours and twenty-five minutes in 1954 to four hours in 1994 and settled at the current three hours thirty minutes in 1997. Even the test developers have changed. ETS developed GMAT questions until 2006 when GMAC changed vendors to ACT Inc. Since 1954 the only things that has not changed is the GMAC’s desire to test those skills necessary to succeed in graduate business schools’ core curriculums, the belief that the skills tested on the exam develop over a relatively long period of time, and quantitative problem solving questions.
Drivers of change in the GMAT
So why has the test changed? In a 1984 paper, 'The Graduate Management Admission Test: Technical Report on Test Development and Score Interpretation for GMAT users', William Schrader suggests that from the beginning the association of business schools that became GMAC recognized that the ATGSB and later the GMAT should change as educational researchers discovered new techniques for testing those skills necessary to succeed in graduate business schools. GMAC largely introduced new questions or modified old question types as a result of testing research. For example, the introduction of Data Sufficiency questions resulted from research indicating that Data Sufficiency questions would increase the reliability of GMAT scores and the introduction of the Analytical Writing Analysis resulted from research indicating that writing ability, as tested by a writing task, was necessary to succeed in most business schools’ MBA programs. Although research has been the impetus for many of the changes to the GMAT exam, GMAC, like a business, has had to control costs while responding to the needs of both its board members and customers.
Many of the board members of GMAC are deans of business schools. As a result, some of the changes to the GMAT exam have resulted from business schools requesting change. In an October 2006 report, 'Use of the GMAT Analytical Writing Assessment: Past and Present', Karen Owens cites surveys that indicate admission committees wanted a writing sample from the AWA to be able to personally validate the AWA score. Additionally, on GMAC’s website under the title 'Why You Can Rely on GMAT Scores', GMAC states the GMAT exam has been repeatedly studied, tested, and modified to ensure that it continues to meet the assessment needs of graduate management programs. GMAC surveyed its board members and responds to their needs.
Why is the GMAT computer adaptive?
There are many theories as to why GMAC chose the Computer Adaptive format. However, when interviewed by BusinessWeek in April 1998, David Wilson the president and CEO of GMAC cited only two reasons for changing to the computer adaptive format: improved customer service and increased flexibility. Prior to the computer adaptive format ETS administered the GMAT only four times a year. Mr. Wilson cited flexibility as being important to MBA applicants who frequently have busy schedules. GMAC understood its customers and responded to their needs.
Other changes to the GMAT exam were made simply for cost effectiveness. For example according to the LinkedIn Profile of Ashok Sarathy, the Director of the GMAT Program at GMAC, the decision to change vendors was the result of a bidding process. Additionally, the introduction of computerized essay readers, ETS’s Erater in 1994 and Vantage Learning’s IntelliMectric in 2006, to score the Analytical Writing Assessment cut down on the cost of employing human readers.
Responsiveness to customer need
By keeping in close contact with board members and sympathizing with customers, GMAC has continued to provide an objective assessment that is relevant to business schools. It is surprising not that the GMAT exam has changed over the last fifty-eight years, but rather how much it has stayed the same. While the format of different questions has changed over time, the GMAC has continued to test skills that develop over a long period and that are necessary to succeed in business schools' core curriculums. Largely the same skills tested in 1954.
Problem Solving questions have not changed since 1954. Data Sufficiency questions have not changed since 1961. It is safe to assume that good quantitative skills have always been necessary to do well in business school. Thus, little change should be anticipated in quantitative question types. However, GMAC has considered making the quantitative section harder. In the April 1998, David Wilson revealed that some business schools would like to see differential equations tested on the exam, implying that the skills taught in some MBA programs require a greater knowledge of mathematics than what is tested. Mr. Wilson suggested that rather than add differential equations to the GMAT exam, GMAC was considering a second test in addition to the GMAT exam. Ten years later the second test has not been introduced. Just as soon as business schools began thinking that more mathematics was required to succeed in business school, they realized what too much mathematics could lead to monumental real world business mistakes. In September 1998 (five months after the BusinessWeek interview), Long Term Capital Management made financial history by turning 2 billon dollars into 6 hundred million in 3 weeks. Nobel Laureate professors at Stanford GSB, the University of Chicago GSB, and Harvard pioneered the sophisticated mathematical models that drove Long Term Capital Management. Two of those professors were partners in Long Term Capital Management. Many business schools took note of the Long Terms demise as evidenced by the preponderance of case studies on the topic. After all comprehending differential equations is not necessary to realize that turning 2 billion dollars into 6 hundred million is not a business savvy move. So perhaps just as GMAC began considering changes to the quantitative portions of the test, the spectacular failure of a hedge fund caused business schools to question whether those changes were necessary. While the quantitative section has remained constant, there have been far more changes to the verbal section.
Emergence of writing and editing skills
In 1954, there was a Sentence Completion, not Sentence Correction, section on the GMAT exam. The questions asked test takers to choose a set of two words to fill in two blank spaces of a sentence. The Sentence Completion questions were first removed from the ATGSB in 1961 and restored in 1966. In 1976, the sentence completion questions were forever discarded in favor of Usage questions. Usage questions asked students to choose which of four words or phrases in a sentence contained an error. If no choice contained an error students were directed to choose the last option, which stated “No Error.” In 1983, Sentence Correction question appeared in their present form. Although there have been small changes to this section of the GMAT exam, there has been a section that tested editing ability, and thus writing ability, on the exam since 1966. However, the portion of the verbal section that is made up of questions that test writing and editing ability has increased significantly. On the first ATGSB, Sentence Completion questions were one of three types of questions tested in a 25 minute section. With the introduction of the Usage section, test takers were given just 15 minutes to complete the Usage section. When Sentence Correction questions were introduced, test takers were asked to complete 22 questions in 25 minutes. The final changes to the length of time allotted for Sentence Correction questions occurred with the introduction of computer adaptive testing. On the computer adaptive GMAT exam roughly half of all verbal questions are Sentence Correction questions. It seems clear that writing and editing abilities have increased in importance since 1954. The trend of increasing emphasis being placed on writing skills is consistent with the introduction of the Analytical Writing Assessment (AWA) in 1994. If the AWA and the Sentence Correction sections are understood both to test writing ability, then approximately 90 minutes of the three and a half hour exam are spent testing writing skills as opposed to approximately 8 minutes on a two hour twenty-five minute exam.
Testing Reasoning Skills
The ATGSB administered in 1954 tested logical reasoning in the Best Arguments section of the exam. Best Argument question consisted of a short passage describing a hypothetical situation followed by questions that asked about the logic of the arguments made in the passage. These questions essentially asked test takers to play judge to resolve a conflict. The Best Argument questions were replaced with Organization of Ideas questions in 1961. Organization of Ideas questions directed test takers to classify statements as central ideas, supporting ideas, illustrative facts, or irrelevant statements. These questions were removed from the GMAT test in 1966. In 1972 Practical Business Judgment questions appeared on the GMAT test in two sections Data Evaluation and Data Application. Data Evaluation questions asked test takers to read a long passage and to classify facts that referenced the passage as major objectives, major factors, minor factors, major assumptions, or unimportant issues. The Data Application section directed students to apply some of the data in the passage to answer question. The Data Application questions were dropped from the GMAT in 1984. Practical Business Judgment questions, which had become known as Analysis of Situation questions, were replaced by Critical Reasoning questions in the late 1980s.
Testing Comprehension Skills
sIn addition to the reading in the Best Arguments section of the first ATGSB test takers reading ability was tested in the Quantitative Reading section. These questions consisted of a passage and possibly a chart. Test takers were asked to infer logical conclusions from the passage. Interestingly, these questions were only used to calculate test takers overall score and were not used to calculate either the verbal or the quantitative sectional scores. Reading Recall or Directed Memory questions replaced Quantitative Reading questions in 1961 and were used in calculating the verbal sectional score. Reading Recall questions directed test takers to read long prose passages and answer questions without referring back to the passage. Preventing test takers from referring back to questions was difficult for test administrators and research suggested that the results of the Reading Recall section wouldn’t change significantly if test takers were permitted to look at the passage. So in 1977, Reading Recall questions were replaced with Reading Comprehension questions.
Analogy and Antonym questions were eliminated from the ATGSB in 1961, reappeared in 1966, and were eliminated again in 1976. They have not returned. Today the only questions on the GMAT that test vocabulary skills associated are Reading Comprehension questions that ask test takers to identify the answer choice that most nearly provides a definition of a word in the passage. It is interesting to note that GMAC made the decision to remove these questions nearly thirty years before they were removed from the SAT in 2005. In 2005 analogy questions were removed due to complaints that the questions were biased and ambiguous. It seems that GMAC’s constant research and improvement has served all stakeholders well and allowed GMAC to anticipate a problem with a question type thirty years in advance.
© Copyright GMATix 2008. All rights reserved
The GMAT® questions, whether taken from the GMAT® mini-test, The Graduate
Management Admission Test: Technical Report on Test Development and Score
Interpretation for GMAT users (1984), or in any other form, are the property of the
Graduate Management Admission Council® and have been reprinted with its permission for
illustrative purposes only in the article titled “History of the GMAT and the associated
GMAT exams - 1954; 1961; 1966; 1972; 1976; 1977; 1984; 1994; and 1997