Lagos IQ Test: What the Test Really Measures

The Lagos IQ Test: What It Really Measures and What It Doesn't

Does an intelligence quotient test measure the same thing in Lagos as in London? A video claiming that more than half of the participants in IQ tests in Lagos (Nigeria) scored below 70 — with a median of 69.7 — is circulating and reopening a debate that science has not settled. The team, according to the version being spread, aimed to refute Western theses about low IQ in African countries. It ended up fueling just the opposite.

The Video That Went Viral

The circulating figures are two: a median of 69.7 and more than half of the participants below 70. From there, some interpretations draw a categorical conclusion about the entire Nigerian population. That leap — from a specific sample to an entire country — is precisely what most methodological analyses reject.

Two things must be separated. One is the data from a specific test in a specific place. Another, very different thing, is what that data allows us to affirm. Between the two lies an abyss that almost no one talks about.

What an IQ Test Really Measures

An IQ test does not measure "intelligence" in the abstract. It measures a limited set of skills: pattern recognition, abstract reasoning, processing estimulante ilegal. Skills that are trainable, that correlate with schooling, and that depend on the type of task one has encountered before. As one of the interventions in the thread summarizes: tests are all trainable and are indirectly representative of academic skills.

Hence the classic example. A Maasai herdsman will get a low score on a matrix test; asking him about tracking an antelope on the savanna is another story. The comparison is not a trick: it's that both exercises measure different things in different contexts. The detail of why the same test works in one place and not another is developed more calmly in the original material.

The Methodological Problems Almost No One Mentions

An IQ test is calibrated on a reference population. If that reference is a country with a consolidated educational system, generalized access to technology, and a stable social structure, the test carries all those assumptions. Applying it to another population without adjusting the control does not measure the same thing: it measures the distance between two environments.

Added to this is the sample design. Data collected in a specific city is not extrapolable to a continent. The inference that converts a median into "more than half below 70" presupposes a normal distribution for which, in such samples, there is no piece-by-piece verification.

The Hereditary Thesis and Its Critics

The uncomfortable part. One current holds that differences in measured IQ between populations have a relevant genetic component, citing adoption studies, twin studies, and comparisons by ancestry. In this vein appear names like Richard Lynn, a collector of IQ data in Nigeria, or James Watson, whose Nobel Prize did not save him from ridicule for his statements.

Against this, the majority stance in population genetics points to something fundamental: the average differences observed between groups in a test do not, by themselves, translate into innate biological differences. The existence of genetic variants with different frequencies among populations — something mentioned in medicine, to the point that certain analyses apply correction factors — does not turn that variation into a ranking of people.

Income and IQ: Correlation Is Not Causation

Maps of economic development and average IQ maps often go hand in hand. What is not resolved is the direction. One hypothesis suggests that a higher economic level improves nutrition, health, and child education, and with it, the measured IQ. Another reverses the order. No one has definitively closed the issue with clean data, partly because the study of the topic itself carries a social stigma that deters researchers.

---

One detail remains that unsettles both sides. If the test measures trainable skills dependent on the environment, the Lagos result speaks above all about the environment that produced it. And if it measures something fixed, then almost no one discussing it daily has the training to interpret it. Comfortable, certainly, for those who were just looking to confirm what they already thought.