What you'll learn

• Why Gould argued that early IQ testing confused intelligence with culture, language and education.
• The AO1 “story” of the study: background, method, results and conclusions.
• How to evaluate Gould using validity, reliability, ethics, sampling bias and ethnocentrism.
• How to compare Gould with the paired contemporary study, Hancock et al. (2011).

Why this study matters

Gould (1982) is the classic core study for the OCR H567 key theme Measuring differences in the Individual Differences area.

The big question is: when psychologists measure a difference between people, are they measuring a real psychological difference — or are they measuring bias in the tool?

Key background concepts

Intelligence and IQ

Intelligence is usually defined as the ability to learn, reason, solve problems and adapt to new situations. However, psychologists disagree about whether intelligence is one general ability or many different abilities.

An IQ test is a standardised test designed to produce a numerical score for intellectual ability. In early testing, this score was often treated as if it measured a fixed, inherited trait.

Mental age

Early intelligence testing used the idea of mental age. For example, if a child performed like the average 10-year-old, they might be said to have a mental age of 10.

This idea becomes much more problematic when applied to adults. Calling an adult “mentally aged 13” can be misleading because adult intelligence, education, language experience and cultural knowledge do not develop in a simple age-line.

Test bias

For Gould, the key issue was that the Army IQ tests used during the First World War claimed to measure innate intelligence, but many items also required English language, schooling, test-taking experience and knowledge of American culture.

Historical background: Binet, Yerkes and eugenics

The original intelligence tests were developed by Alfred Binet in France to identify children who needed educational support. Binet did not intend IQ to be used as a fixed label of worth.

In the USA, psychologists such as Robert Yerkes adapted intelligence testing for large-scale use. During the First World War, Yerkes and colleagues tested huge numbers of army recruits using group intelligence tests.

This happened in a wider social context of eugenics.

Gould argued that IQ testing was used to support racist and anti-immigration views, especially the idea that some ethnic groups were naturally less intelligent.

The study story: Gould (1982)

Aim

Gould aimed to challenge the claim that the US Army intelligence tests measured innate intelligence fairly. He wanted to show that the tests were biased and that the results had been misused to support discriminatory conclusions.

Design

Gould’s work was not a laboratory experiment. It was a historical review and secondary data analysis.

Gould examined Yerkes’ published army testing data, test materials, administration procedures and the conclusions drawn from the results.

Sample and source material

The original data came from around 1.75 million US Army recruits tested during 1917–1918. They were male military recruits, including many immigrants and men with different levels of education and English language ability.

Gould’s “sample” was therefore not a new group of participants. It was the existing historical record of Army Alpha and Army Beta testing.

Materials and apparatus

The main testing materials were:

• Army Alpha: a written verbal test for literate English-speaking recruits.
• Army Beta: a pictorial or non-verbal test intended for illiterate recruits or those who did not speak English well.
• Individual examinations: used for some recruits who failed earlier testing.

Procedure

Recruits were usually given Army Alpha if they were judged literate in English. Those who failed Alpha, or who were considered illiterate or non-English-speaking, could be moved to Army Beta. Some who failed Beta received individual assessment.

Gould reviewed the content of the tests and the testing conditions. He considered whether low scores could be explained by factors such as poor English, limited schooling, unfamiliarity with American culture, rushed group testing and anxiety in a military setting.

Results

Yerkes’ interpretation of the results was alarming: the average white American recruit was said to have a mental age of about 13. This supported the famous phrase “a nation of morons”.

The tests also appeared to show that some immigrant groups scored lower than others, and that Black recruits scored particularly low. Hereditarian psychologists used this to argue that intelligence differences were inherited and linked to race or nationality.

Gould disagreed. He argued that the pattern made much more sense if the scores reflected acculturation, education and English language exposure. Immigrants who had lived in the USA longer tended to do better, which suggests the test measured familiarity with American culture rather than fixed innate intelligence.

Conclusions

Gould concluded that the Army tests lacked validity as measures of innate intelligence. They were affected by language, culture, schooling and administration conditions.

He also argued that the results were socially dangerous because they were used to justify eugenics and immigration restriction.

Evaluating Gould: AO3

Validity

The key evaluation point is construct validity.

Yerkes claimed the tests measured innate intelligence. Gould argued they actually measured a mixture of intelligence, English literacy, cultural knowledge, education and confidence with formal tests.

This is a strong criticism because if the measure is invalid, the impressive number of participants does not rescue the conclusion.

Reliability

The Army tests were standardised in the sense that many recruits received similar materials. However, testing conditions were not always consistent. Recruits could be tested in large groups, under time pressure, in unfamiliar military settings.

Gould’s own review is also open to the criticism of researcher bias. He selected and interpreted historical evidence to support an argument. However, because he used published materials and quoted test items, other researchers can inspect the same sources.

Sampling bias and generalisability

The original Army sample was large, but not representative of the whole population. It included military-age men, not women, children, older adults or people outside the recruitment system.

There were also major differences in education, social class, immigration history and language background. Gould’s point is that these differences were treated as evidence of innate intelligence when they may have reflected social opportunity.

Ethics and socially sensitive research

The original Army testing raises serious ethical issues. Recruits were in a military hierarchy, so genuine consent and right to withdraw were limited. The results labelled groups as inferior and contributed to discriminatory policies.

Under the modern BPS Code of Human Research Ethics, psychologists should consider respect, responsibility, integrity and social impact. Gould’s study is important because it exposes how psychological measurement can harm people when used carelessly.

Applying Gould to a new situation: AO2

Gould is very useful for applying to modern examples, such as schools, workplaces or immigration systems using ability tests.

Comparison with Hancock et al. (2011)

The paired contemporary study is Hancock et al. (2011), often known as Hungry like the wolf. It investigated whether psychopathic murderers used language differently from non-psychopathic murderers.

Both studies show that measuring individual differences can be useful, but also risky. Gould shows how measurement can be distorted by culture and politics. Hancock et al. used more modern, systematic tools, but still raises questions about validity, sampling and socially sensitive labels.

Links to the Individual Differences area

The Individual Differences area studies how people vary, for example in intelligence, personality, mental health or criminal behaviour.

Gould fits this area because he focuses on whether psychologists can fairly measure differences between people. His study also links strongly to debates:

• Nature/nurture: Yerkes emphasised inherited intelligence; Gould emphasised environment, education and culture.
• Reductionism: IQ testing reduced complex human ability to a single score.
• Ethnocentrism: the tests favoured American English-speaking cultural knowledge.
• Usefulness: Gould’s critique helps improve fair testing, but the original testing was harmful.
• Psychology as a science: numbers and tests can look objective, but scientific measurement depends on valid concepts and fair methods.