Are IQ Tests Biased Against Africans?
Conclusive evidence suggests there is a ten to fifteen-point gap in IQ between whites and African Americans in the United States. In the same country, there is a five-point IQ gap between Asians and whites. These results have been consistent globally from as early as the 1920s with East Asians averaging an IQ of 108, whites averaging an IQ of 103, and African Americans averaging an IQ of 91. There was a significant narrowing of IQ gaps in the 60s and 70s, but the narrowing has since stopped and remained largely the same for the last three decades. Many people, especially Africans, have objected to these findings, calling them white supremacist gibberish. Since the data presented above is mostly of the United States, it’s worth mentioning that Sub-Saharan Africa has an average IQ of 68.9. This average is well below the African American mean as well as the white and Asian means.
Nothing is as controversial as the means presented above. When I was growing up, it was controversial for evolution to be true. Today, the evolution of our bodies is no more controversial than our evolution above the shoulders. People are willing to accept that evolution is true, if and only if, it precludes group differences in cognitive ability. Recently, Professor John Sherman a chemistry lecturer at the University of British Columbia was suspended for basically relaying the same facts I’ve presented above. Sherman said that:
Well, I’ll give you a fact. Black people do poorer on IQ tests than white people. Is that racist? That’s a fact. Is it racist to say it? I don’t know. Asians do better than white people on IQ tests, and that’s a fact. Now, you can make judgements from that. Does that mean Asians are smarter than whites, and whites are smarter than blacks? I don’t know. And is that racist to even ask that question? I don’t think it’s racist.
Generally, Professor Sherman was relaying facts that have been true from as early as the 1920s. The fact that he was cancelled for stating the obvious tells us just how much talking about intelligence is akin to walking on eggshells. It doesn’t hurt to reveal that I am an African. I’ll also confess and say that over the past few years I’ve caught some strays too. I get called a white supremacist and a racist all the time. However, the point of this essay is not to discuss my woes. Instead, I am looking to present arguments that show why IQ tests are not biased against Africans in any way. Actually, most people will be surprised to learn that measures of IQ are more valid, reliable, and credible than the pop science most of us learn in psychology. These fields have a major replication problem, partly driven by the desire to get positive results that don’t hurt anyone’s feelings. However, intelligence research as a mostly statistical field, has enjoyed large piles of evidence and massive replication for generations now.
First, the idea IQ tests are racist needs to be laid to rest now. The same IQ tests that reveal massive gaps between Africans and whites also reveal massive gaps between Asians and whites. Is it racist that Asians score higher than whites in IQ tests? If your answer is no, then why would it be racist that whites score higher than blacks in the same tests? If your answer is yes; that IQ tests are indeed racist, then why would white supremacists build IQ tests that glorify Asian supremacy? As you’ve noted, these are similar questions that professor Sherman asks regarding whether IQ tests are racist products of white supremacy. The idea that a test is racist only when it reveals disparities between whites and blacks but not when it shows the gaps between Asians and whites sounds like cherry peaking to me.
Objections against IQ tests and the fact that they reveal group differences is largely motivated by various political positions on the progressive left, which decry the existence of innate human abilities. To them, groups as well as individuals are equal, and that any disparities between them can only be a product of systemic marginalization, oppression, and exclusion. Steven Pinker argues that human nature and the inherent differences between people and groups can only be made worse by our denial of their existence. He also opines that just because differences between groups exist does not imply that discrimination and racism is justified. As moral beings, it’s possible to co-exist despite the inherent differences between us. The same way we differ in our height, weight, and other factors is the same way we differ in intelligence. Denying these differences, or even worse, attributing them to racism and discrimination without considering other genetic and environmental factors is lazy. It’s also worth noting that IQ tests are measurement tools, and while they may reveal differences, they do not tell us the source of those differences. Therefore, while we can sit and debate the causes of group differences in cognitive ability, it’s futile to sit and debate whether these differences exist in the first place.
Africans and African Americans have presented several objections to IQ tests and the findings of IQ research. According to critics, IQ tests are racially and culturally biased in favor of white people. Arguments presented posit that most IQ tests reflect western middle class cultures which Africans and African Americans may not be conversant with. In most cases, the use of the English language in various test batteries is a reflection of the cultural bias baked into those tests. However, expert opinion in the field of intelligence research and psychometrics is pretty much settled. Charlie Reeve and Jennifer Charles conducted a survey of opinions, where they asked intelligence experts about their views regarding various findings on IQ. There was a consensus across-board with most of them agreeing with claims such as; “General Cognitive Ability is measured reasonably well by standardized tests.” These experts agree that indeed IQ tests are a proper measure of intelligence. Experts also agreed that there was no “valid substitute of g,” meaning there is no other way to explain human cognitive ability. While not a rebuttal, their consensus lays to rest the theory of multiple intelligences by Howard Gardner, Carol Dweck’s mindset theory, and even Robert Sternberg’s theory of practical intelligence.
Experts also agreed that, “general cognitive ability tests are fair,” and that “professionally developed CATs are not biased against members of…minority groups.” Since experts already agree that IQ tests are indeed fair and that they do not discriminate against Africans or any other minority group, then we are compelled to accept their views, unless all of them are racist white supremacists. However, I won’t ask you to “trust the science” just yet. If what we’ve seen during the Covid-19 pandemic is to go by, then we can agree that experts can pretty much be influenced by their politics as well as individual prejudices. It is important to look at the specific ways IQ tests are said to discriminate Africans then disprove them.
Critics observe that IQ tests are culturally biased in various test items. Russel Warne, for example, observes that one test item that has attracted a lot of criticism is the “fight item” which appeared in a now decommissioned version of WISC. The test item asked; “what is the thing to do when a fellow much smaller than you starts a fight?” Critics believe this item was culturally biased against some cultures where aggressiveness is valued. A good example is African American cultures which tend to have higher rates of crime and violence. A child raised in such a culture may as well think it’s okay to fight. However, a closer analysis of the test item revealed African Americans did not perform any worse than Europeans. It’s a tendency for critics, however, to rely on items like the “fight item” to discredit entire IQ tests. The correct answers in the fight item was to either run away or tell an adult. Any student that picked “fight back” failed that item.
The same “fight item was adopted by a black Zimbabwean psychologist called Zandi who administered it to Zimbabwean children. Findings from her results revealed that the test item was not culturally inappropriate for black Zimbabwean children. While some items in an IQ test might attract considerable criticism, it’s never readily apparent to any individual whether any of the items in a test are culturally inappropriate. As an individual, you cannot read a test item and conclude it’s culturally inappropriate; what if previous performances on the item have not shown any disparate performances across cultures? I believe the same is true of most IQ tests administered in English. While many people might believe that an IQ test in English disadvantages some non-native speakers, it is pretty much common for some none-native speakers to excel in those tests. Many African schools, for example, teach English in grade school and high school, and are expected to out-compete Europeans in verbal tests. However, to remove any doubt, researchers typically measure the IQs of none-native English speakers with IQ tests specifically developed for those cultures, or they rely on non-verbal tests.
Non-verbal tests, like the one presented above, include questions with matrix items or digit spans. Matrix items comprise a bunch of shapes and figures while digit spans comprise a string of numbers to be repeated in order. Typically, nobody can claim that a bunch of shapes or numbers are culturally inappropriate, since numbers and shapes are universal, and only require us to look and infer. In many studies conducted in Kenya and other African countries, the Raven’s Progressive Matrices and the Raven’s Colored Progressive Matrices are used to measure IQs. The RCPM is a culturally unbiased test comprising mostly of shapes meant to measure spatial and reasoning ability. Unlike the verbal tests, these tests are highly g-loaded and exert significant demands on the examinees’ cognitive faculties. Robert Sternberg looking to test the practical intelligence of young boys and girls in Western Kenya relied on various verbal and non-verbal tests to measure IQ. Among the non-verbal test was Raven’s Colored Progressive Matrices. The author also used the English Mill Hill Vocabulary Scale, and a Dholuo Vocabulary Scale. Notice the Dholuo test is a local adaptation of a culturally appropriate IQ test. This is how Sternberg defines the Dholuo Vocabulary Scale:
This test was designed to measure crystallized intelligence based on the language spoken at home. It consisted of 33 items. Words were in Dholuo and children were asked to write a synonym next to each target word.
Charlotte Neumann and colleagues also measured the IQs of Kenyan students in Embu. Most of the tests adopted for use in the study were non-verbal, including the culturally appropriate Ravens Progressive Matrices which measures reasoning by analogy, problem solving, abstract reasoning, and perceptual awareness. The authors also used the Verbal Meaning Test which they claim was “designed in East Africa, uses pictures of familiar objects and people and is used to measure receptive language and vocabulary.” The researchers were careful not to rely on any tests that would invite criticisms such as is the case with most English verbal tests.
A different study led by Patricia Kitsao-Wekulo was conducted in the Kenyan Coast of Kilifi. The researchers in this study did several pilot studies with the children, ensuring the test items used were culturally appropriate. The pilots lasted for around 90 minutes each and the children were provided with “appropriate translations of materials.” As the researchers suggest:
The specific goals of piloting were to make modifications that avoided floor and ceiling effects, facilitated the child’s engagement and understanding, ensured ease in test administration and scoring, and demonstrated face validity as established through interviews with children and adults.
The study also relied on local assessors, each fluent in the testing languages. Among other tests, the Ravens Progressive Matrices, the Tower Test, Self-Ordered Pointing Test, and Verbal List Learning tests were used.
Most IQ tests used in Kenya and Africa are generally designed as outlined above. There’s considerable rigor when designing culturally appropriate tests. Specifically, non-verbal IQ tests such as Raven’s Progressive Matrices are used. In instances where verbal tests are essential, either a culturally appropriate verbal test such as the Dholuo Vocabulary Scale is used or the examinees are primed earlier with relevant translations and materials, and the tests themselves are administered by local assessors. This format also dispels the argument that most African children are either indifferent to IQ tests administered by white researchers, or they are reluctant to participate in the tests; attitudes believed to subsequently affects their scores. Russell Warne, for example, highlights the Panga Munthu Test which was developed in Zambia. He says that:
Panga Munthu’s creators believed that intelligence is universal…instead of responding to questions verbally or using pencil and paper (often unavailable in rural Zambia), the Panga Munthu requires children to sculpt a human figure in clay or wire –a common activity for children in Zambia.
IQ tests, therefore, don’t have to be administered in English as many people falsely believe. IQ tests can be adopted for local use by relying on local languages and practices to present test items. Considering that most IQ tests are also administered by local teachers and assessors, it is also unlikely that students would be unwilling to participate in the tests.
After seeing how IQ tests are deigned and administered, the idea that test gaps are a result of discrimination, racism, or white supremacy quickly becomes preposterous. Most studies conducted in Kenya actually aimed at accomplishing social good. For example, the study I have hinted at above by Charlotte Neumann aimed to find whether adding meat supplements to the food eaten by Kenyan school children in marginalized regions would improve their cognitive ability. Another study led by Constance Gewa in Embu tested the effect of various dietary micronutrients to the cognitive performance of children. A different study in Embu led by Tamara Daley was first to establish the Flynn effect in Kenya, and is one of the studies that gives us hope that IQs in Africa will rise as time passes by. None of these researchers seem to have any particular ill motives. Suffice to note that some of these researchers are Africans.
IQ tests have been criticized for being culturally biased. However, in light of the evidence presented here, those ideas seem preposterous. Expert opinion by intelligence researchers and psychometricians suggests that test items are fair and not biased towards any minority groups. Test items such as the “fight item” that was presumed to be culturally biased did not show any bias when used across different cultures. In other cases, researchers relied on non-verbal tests such as the Raven’s progressive matrices to measure the IQ of children. These tests do not have any cultural bias baked into them unlike the verbal tests typically presented in English. Furthermore, in instances where verbal tests were essential, customized verbal tests specific to given cultures were used. In Western Kenya, the Dholuo Vocabulary Scale was used to measure crystalized intelligence. The Panga Munthu Test was designed for use in Zambia, and presented another culturally appropriate way to administer IQ tests. Therefore, there is very little room for cultural as well as racial bias in IQ testing and psychometrics in Africa. The same is true of IQ tests administered elsewhere including the United States.
