When Analogies Ruled the SAT

• Introduction
• Chapter 1 The Army Alpha and the Dawn of Standardized Logic
• Chapter 2 Carl Brigham’s Big Idea: Measuring Pure Aptitude
• Chapter 3 A is to B: The Mechanics of the Relational Pair
• Chapter 4 The Ivy League Gatekeeper: ETS Takes the Reins
• Chapter 5 The Distractor Factor: The Art of Crafting Wrong Answers
• Chapter 6 OSTRICH : BIRD :: The Architecture of the 1950s SAT
• Chapter 7 The Vocabulary Trap: Words You Only Need Once
• Chapter 8 Stanley Kaplan and the First Test Prep Guerrillas
• Chapter 9 Class Markers and Cultural Codes: The Demographics of Diction
• Chapter 10 The Infamous REGATTA : CABIN Controversy
• Chapter 11 High Stakes, High Stress: Student Narratives from the Golden Era
• Chapter 12 What Cognitive Science Thought Analogies Measured
• Chapter 13 The Psychometric Wars: Aptitude versus Achievement
• Chapter 14 The 1990s Revision: The Analogy Under Siege
• Chapter 15 Decoding the Secret Code: The Rise of Analogy Shortcuts
• Chapter 16 The University of California Threatens to Walk
• Chapter 17 Richard Atkinson and the Battle for the SAT’s Soul
• Chapter 18 The Final Verdict: Why the Analogy Had to Die
• Chapter 19 March 2005: The Last Pair is Scored
• Chapter 20 The Rise of Critical Reading and Sentence Completion
• Chapter 21 The Ghost of Analogies in Modern Cognitive Testing
• Chapter 22 What We Lost When We Lost the Analogy
• Chapter 23 The New Gatekeepers: How Testing Evolved Post-2005
• Chapter 24 The Legacy of America's Most Hated Question Type
• Chapter 25 The Future of Measuring Human Intelligence

For five decades, millions of American high school students faced down a very specific form of academic torture. Sitting in dim high school cafeterias and gymnasium rows on Saturday mornings, armed with wooden No. 2 pencils, teenagers were presented with cryptic strings of capitalized words separated by colons and double colons. RUNNER : MARATHON. OSTRICH : BIRD. REGATTA : CABIN. To pass through the gates of higher education, one had to master the cryptic syntax of relational pairs, discerning subtle differences between degrees of intensity, functional relationships, and category memberships. The SAT analogy was more than just a test question; it was the ultimate symbol of American meritocracy, a gatekeeper that promised to measure raw, unadulterated human aptitude regardless of high school curriculum or background.

Yet behind the austere front of psychometric objectivity lay a question type that was intensely polarizing, deeply flawed, and uniquely stressful. For fifty years, from its rise out of World War I military testing to its dramatic execution in March 2005, the verbal analogy ruled the SAT. It was built on a bold premise: that by stripping away context and focusing entirely on abstract relationships, test designers could measure a student's pure ability to reason—their innate cognitive muscle. But what began as an idealistic attempt to find diamond-in-the-rough talent across a vast nation quickly evolved into a high-stakes battleground involving cultural politics, test-prep guerrillas, Ivy League gatekeepers, and the very definition of intelligence itself.

This book tells the definitive story of the analogy’s half-century reign: how it was conceived, how it functioned, why it was defended with almost religious zeal by psychometricians, and why it ultimately had to die. It traces the arc of an American obsession with standardized measurement, beginning with the early intelligence tests of the 1920s and moving through the mid-century rise of the Educational Testing Service. Along the way, it explores the intricate art of distractor construction—the science of building wrong answers so deviously plausible that they shattered the dreams of ambitious applicants—and examines how an item designed to level the playing field inadvertently became a proxy for socioeconomic class and cultural exposure.

When Analogies Ruled the SAT is not merely a nostalgic autopsy of a defunct test section; it is a cultural and scientific biography of a national trial by fire. It brings readers into the quiet war rooms of psychometricians, the high-stress test centers of the late 20th century, and the executive suites of the University of California system, whose public threat to abandon the SAT entirely delivered the death blow to the analogy pair. Through student narratives, archival documents, cognitive research, and historical analysis, this book illuminates the complex web of forces—psychometric, political, and economic—that dictated how generations of Americans were sorted, ranked, and selected.

By understanding the life and death of the analogy question, we gain a clearer lens through which to view today's ongoing debates over standardized testing, merit, and educational access. When the last analogy was scored in the spring of 2005, it marked the end of an era in American assessment—the moment we abandoned the dream of measuring pure, isolated "aptitude" in favor of curricular achievement and contextual reading. Whether that transition represented progress, a surrender to commercial pressures, or the loss of a valuable cognitive tool remains one of the most compelling questions in modern education. This is the story of how two words, separated by two dots, came to rule American education—and why we are still living in the aftermath of their collapse.

In the spring of 1917, the United States found itself facing a crisis of logistical management that had nothing to do with iron ore, grain production, or ammunition manufacturing. President Woodrow Wilson had declared war on Germany, and the American military, which stood at a modest strength of roughly 130,000 men, was tasked with expanding into a fighting force of over two million within months. The Selective Service Act flooded military encampments with a chaotic cross-section of American manhood: farmhands from the Midwest who had never seen a city, recent immigrants from Southern and Eastern Europe who spoke minimal English, factory workers, coal miners, and fresh-faced Ivy League graduates.

The War Department faced an unprecedented organizational nightmare. Officers needed to identify, within weeks, who possessed the raw mental capacity to operate heavy artillery, navigate complex logistics, lead platoons, or manage signal communications—and who was suited only for manual labor. Traditional methods of evaluation, such as face-to-face interviews or commander recommendations, were hopelessly slow and notoriously subjective. Into this administrative void stepped an ambitious, forty-one-year-old Harvard psychologist named Robert Mearns Yerkes.

Yerkes, then president of the American Psychological Association, recognized a historical opportunity. For decades, academic psychology had struggled to establish itself as a hard science alongside physics and chemistry. Skeptics viewed experimental psychology as little more than philosophical speculation adorned with brass instruments and stopwatches. Yerkes believed that if psychologists could demonstrate their practical utility by scientifically sorting the United States Army, the discipline would forever secure its status as a recognized authority in public life.

In May 1917, Yerkes assembled a committee of seven leading psychological researchers at the Vineland Training School in New Jersey. The group included Lewis Terman, the developer of the Stanford-Binet intelligence scale, and Henry Goddard, a prominent intelligence researcher. Their goal was simple yet audacious: construct a standardized, group-administrable paper-and-pencil test capable of measuring the raw intelligence of hundreds of thousands of men simultaneously.

Before this gathering, intelligence testing had been a painstaking, one-on-one affair. Alfred Binet’s original French test, designed at the turn of the century to identify school children needing special instruction, required a trained clinician to sit across a table from a single subject, asking open-ended questions and timing their responses. Such a method was entirely unworkable for an army mobilizing millions.

The breakthrough came via Arthur Sinton Otis, a graduate student of Terman’s who had spent the preceding years developing a method for mass group testing. Otis introduced the revolutionary concept of multiple-choice and short-answer matching questions. By converting open-ended psychological evaluations into objective items with a single correct answer pre-printed on a key, Otis removed the need for clinical judgment. Anyone capable of holding a timer and checking marks on paper could score the test.

The committee spent weeks drafting, debating, and refining test items. The result of their labor was a battery of examinations that would transform American institutional history: the Army Alpha for English-literate recruits, and the Army Beta, a non-verbal pictorial test for illiterate soldiers and those non-fluent in English. The modern industry of mass standardized testing was born.

Inside the Army Alpha

The Army Alpha consisted of eight distinct subtests, designed to measure various facets of cognitive functioning under tight time constraints. Recruits sat shoulder-to-shoulder on wooden benches in drafty cantonments, balance boards resting on their knees, gripping stubby lead pencils. A psychometrically trained lieutenant would stand at the front of the room, shouting precise instructions from a standardized manual while keeping time with a stopwatch.

The subtests moved at a relentless pace. Test 1 evaluated the ability to follow oral commands, requiring recruits to carry out complex geometric instructions on paper within seconds ("When I say 'go,' make a cross in the second circle and a line through the third"). Test 2 presented arithmetic problems. Test 3 tested practical judgment, asking questions such as why street cars use electric power instead of horses. Test 4 focused on synonym and antonym recognition. Test 5 tested disarranged sentences. Test 6 involved number-series completion. Test 8 measured general information, asking recruits to identify the manufacturer of the Buick automobile or name the city where the National League baseball team played.

It was Test 7, however, that introduced the broad American public to a novel exercise in abstract relation matching: the verbal analogy.

Test 7 was titled simply "Analogies." It presented recruits with forty items to be completed in just three minutes—a grueling pace of four and a half seconds per question. Each item featured three given words followed by four choice options. The instructions printed at the top of the sheet explained the logic with stark clarity: "In each of the lines below, the first two words are related to each other in some way. See what that relation is, and then find the word in the bracketed four words that is related in the same way to the third word."

Item 1 set the template: dog - bark :: cat - (mew, cry, roar, chirp)

To modern eyes accustomed to the high-level vocabulary of mid-century college entrance exams, the Army Alpha’s analogies seem deceptively simple. Early items featured everyday domestic pairs: shoe - foot :: hat - (coat, nose, head, collar) water - drink :: bread - (eat, cake, hungry, bake) day - night :: white - (red, black, clear, pure)

Yet as recruits progressed down the page, the logical relationships became progressively more abstract, demands on working memory increased, and the vocabulary sharpened: granary - wheat :: library - (books, desk, read, building) abate - lessen :: augment - (increase, divide, subtract, multiply) statue - shape :: song - (tune, words, instrument, beauty)

The inclusion of the analogy subtest was not an accident; it reflected a deliberate theoretical commitment by early psychometricians. Psychologists like Terman and Yerkes were heavily influenced by Charles Spearman, the British statistician who had proposed the theory of g—a single, general intelligence factor underlying all mental performance. Spearman argued that human intelligence was defined by two primary operations: the apprehension of experience and the "eduction of relations and correlates."

In plain English, Spearman believed that true mental ability was not the mere retention of memorized facts, but the capacity to perceive the underlying structural relationship between two ideas and immediately apply that exact relationship to an entirely new domain. The verbal analogy was prized precisely because it was viewed as the purest possible operational test of this ability. It stripped away narrative context, rhetorical flourish, and long-form exposition, isolating the core logical engine of the human mind.

The Delusion of Pure Aptitude

From its inception, the logic of the verbal analogy rested on a fundamental assumption: that relational thinking was an intrinsic biological capacity, largely independent of schooling, wealth, or cultural background. Test creators believed that an uneducated farm hand with high innate intelligence would intuitively recognize that a granary stores wheat just as a library stores books, even if he had never set foot inside a grand municipal library.

In practice, however, the early Army Alpha analogies were heavily saturated with specific cultural, regional, and socioeconomic assumptions. To solve an analogy, a candidate must first know the precise definitions of all four terms in the prompt; if the vocabulary is unfamiliar, the underlying relational logic becomes entirely inaccessible, no matter how sharp the test-taker's cognitive machinery might be.

Consider an item from the 1917 testing battery: tear - sorrow :: smile - (joy, face, lips, laugh)

While this item relies on universal human emotions, other items on the Army Alpha demanded familiarity with specific middle-class American lifestyle details of the early twentieth century. Recruits were asked to map relationships involving types of fabrics, specialized farming tools, commercial products, and institutional structures that favored native-born, urban, educated Americans. A recruit who grew up speaking Yiddish in Lower Manhattan or Italian in rural Pennsylvania might possess exceptional analytical capacity, but when confronted with words like harrow, silo, or cleaver, the logical framework collapsed instantly.

Furthermore, the environment in which these tests were administered was far from controlled. Hundreds of thousands of draftees were marched into overcrowded barracks, often after days of exhausting physical drill, and forced to complete the test under strict time limits that penalized deliberation. For many recruits, it was the first time they had ever held a pencil or looked at a standardized printed grid.

Despite these manifest distortions, Yerkes and his colleagues viewed the results of the Army testing program as an unmitigated triumph for psychological science. Between 1917 and 1918, the War Department administered the Army Alpha and Beta tests to over 1.7 million men. It stood as the largest psychological experiment in human history.

Psychologists compiled massive statistical tables categorizing soldiers into intelligence grades from A to E. Men who scored high marks were routinely flagged for Officer Candidate School or specialized technical units, while those scoring poorly were assigned to infantry labor battalions or discharged entirely as "mentally deficient." For the first time, an individual's career trajectory within a massive national institution was determined by his performance on a timed, paper-and-pencil test of abstract logic.

The Great Post-War Shockwave

When the armistice was signed in November 1918, the military rapidly demobilized, but the data collected by Yerkes’ team remained. In 1921, the National Academy of Sciences published Yerkes' massive, 890-page report titled Psychological Examining in the United States Army. The findings sent a profound shockwave through American intellectual and political culture.

The most widely publicized and controversial statistic to emerge from the data was the assertion that the average mental age of white American adult male recruits was just 13.08 years. Immigrants from Southern and Eastern Europe scored lower still, with average mental ages hovering between 10 and 11 years. African American recruits from Southern states registered the lowest average scores on the scale.

The public reaction was immediate and intense. Alarmists and popular commentators latched onto the figures as proof that the American populace was undergoing a process of rapid intellectual degeneration, fueling racist anti-immigration arguments that culminated in the restrictive Emergency Quota Act of 1921 and the Immigration Act of 1924. Eugenicists pointed to the Army Alpha data as scientific validation for their claims that intelligence was strictly hereditary and that the nation's genetic stock was diluting.

A few critics saw the obvious flaws in the methodology. Journalist Walter Lippmann launched a scathing series of articles in The New Republic in 1922, attacking Yerkes and Terman. Lippmann argued that the psychologists were guilty of extraordinary intellectual arrogance, mistaking acquired language skills, schooling, and familiarity with American culture for innate, biological mental capacity. He pointed out that a test measuring whether a man knew that butter was made from milk or that a Buick was a car was measuring exposure to specific life experiences, not pure intelligence.

Terman fired back in public debates, defending the integrity of the tests and insisting that the paper-and-pencil instruments measured fundamental cognitive abilities that high schools and colleges could not alter. The debate was fierce, but as the 1920s progressed, the institutionally minded proponents of testing won the war of practical adoption.

The wartime experiment had proved something of immense commercial value: mass standardized testing was cheap, scalable, efficient, and bestowed a powerful aura of scientific authority on institutional decision-making. Businesses began using modified versions of the Army Alpha to screen job applicants. Municipalities used them to hire police officers and firefighters. Most importantly, school administrators saw in the Army Alpha a model for organizing public education.

From the Barracks to the Classroom

Prior to the 1920s, American school admissions and placements were localized, informal, and heavily reliant on essay examinations graded by individual instructors. The influx of high school students following World War I overwhelmed traditional systems. School boards desperately needed an efficient mechanism to track students into vocational, commercial, or academic streams.

Arthur Otis, the man who had invented the multiple-choice format for the Army Alpha, translated his wartime research directly into the private sector, publishing the Otis Group Intelligence Scale. Lewis Terman published the National Intelligence Tests for elementary school children. Within a few short years, millions of American school children were sitting at wooden desks, filling out printed booklets containing number series, synonym scales, and the now-familiar three-word analogy chains.

The analogy had successfully migrated from the draft board to the classroom. It was celebrated by educators for its unique structural features. Unlike essay questions, which took hours to grade and invited subjective bias, an analogy item had a single, incontrovertible answer that could be checked in seconds using an overlay stencil. Unlike simple vocabulary tests, which merely measured rote memorization, the analogy required a secondary mental step: the student had to perform an operational maneuver, identifying the precise abstract rule connecting the first pair of words and holding that rule in working memory while searching for its duplicate among the answer choices.

It was this double-action mechanics—combining lexical knowledge with relational logic—that made the analogy format irresistibly attractive to psychometricians. It seemed to offer the holy grail of educational measurement: a concise, elegant, bulletproof tool for separating students who merely memorized facts from those who possessed real analytical capability.

Among the young scholars who observed the wartime success of the Army Alpha with keen interest was a bright, mathematically inclined assistant professor of psychology at Princeton University named Carl Campbell Brigham. Brigham had served directly under Robert Yerkes in the Sanitary Corps during the war, helping analyze the mountains of data generated by the Army testing program.

Brigham was convinced that the methods refined in the military barracks could be applied to solve a growing crisis in American higher education: how to select candidates for university admission out of a vast, highly unequal, and rapidly expanding high school population. Armed with the datasets of the Army Alpha and an unshakeable faith in the power of statistical analysis, Brigham sat down to design a new exam specifically tailored for college applicants. When he did, he made sure that the verbal analogy was positioned at the very center of his blueprint.