The Dark History of IQ Testing: From the 1917 Army to Eugenics

The first large-scale intelligence test did not begin as a laboratory curiosity. It began as a military tool. In 1917, when the United States entered the First World War, psychologist Robert Yerkes led a project to evaluate Army recruits. About 1.7 million men were tested. What followed is one of the more uncomfortable lessons, and one of the more necessary, in how a number can be abused.

Army Alpha and Army Beta

There were two group tests. The Alpha was written, for men who could read English. The Beta was pictorial, built for illiterate recruits and for men who did not speak the language. The stated goal was to place each recruit in the job that best fit his abilities and to spot who might make an officer.

The trouble was the conditions. The tests were given in packed rooms, with instructions shouted or acted out, under tight time limits, and many examinees were unfamiliar with pencil and paper or with the cultural vocabulary the questions took for granted. Even so, the results were published as if they cleanly measured innate ability.

From test scores to immigration quotas

Here comes the gravest turn. Carl Brigham, a psychologist who had worked with the Army data, published A Study of American Intelligence in 1923, arguing that immigrants from southern and eastern Europe were less intelligent than those from northern Europe. Those conclusions overlooked decisive variables: command of English, years of schooling and poverty. They circulated anyway, with enormous authority, in public debate.

In 1924 Congress passed the Johnson-Reed Immigration Act, which set quotas by nationality and favored northern European countries. Historians debate how much weight the tests carried compared with the racism and politics of the day, but they agree the tests gave a scientific gloss to arguments that already existed. Brigham publicly recanted in 1930, admitting his conclusions had no foundation. Only a few years earlier he had designed the first SAT, an irony history did not spare us.

Eugenics and forced sterilization

Eugenics was a movement that aimed to "improve" the population by controlling who had children. Henry Goddard, who translated Binet's scale into English and popularized the term "moron" for people with low scores, argued that such people should not reproduce. Binet, by contrast, had warned that his scale should never be used to label a child permanently.

In 1927, in Buck v. Bell, the U.S. Supreme Court upheld the forced sterilization of Carrie Buck, a young Virginia woman classified as "feeble-minded." Justice Oliver Wendell Holmes wrote a line that became infamous: "Three generations of imbeciles are enough." Later research showed the case rested on very flimsy evidence about Carrie and her family. Over the course of the twentieth century, tens of thousands of people were sterilized in the United States under laws of this kind.

Why it took so long to correct

In part because numbers command respect. A score in a table looks objective even when it was gathered in poor conditions from a badly defined sample. The conclusions also matched prejudices already in circulation, and people review what confirms their beliefs less carefully. Over the years, other researchers pointed out the flaws and the standards changed. Today a test is expected to document how it was built, who it was given to when setting norms, and what evidence shows it measures what it claims to measure.

Knowing this chapter makes us better readers of any statistic. Asking who measured, how, on whom and for what purpose remains a good habit, whether the subject is an IQ test or a news headline.

What is left of the lesson

The lesson is not that IQ tests are useless. Modern tests are built with strict standardization, bias review and individual administration. The lesson is about how they get used:

• A score describes performance on one specific test, not the worth of a person or a group.
• Comparing groups without accounting for education, language and living conditions leads to false conclusions.
• A number should never be used to deny rights or to decide who deserves to have children or to immigrate.

That is why we keep saying at TestIQ that your result is an orientative snapshot of how you solve certain problems today, nothing more. If you would like to try it in that spirit, the test is waiting.