Monday, March 15, 2021

Nigerians, Scrabble, and the GCSE


Exam hall at Hull Collegiate School (Wikicommons – Robin S. Taylor). The GCSE exam is a poor measure of raw cognitive ability. If some students get tutored and others do not, there will be more environmental variance in IQ, and the exam results will say less about the genetic potential for cognitive ability.



Chanda Chisala has written more about cognitive ability in sub-Saharan Africa. His argument is straightforward:


[…] if it is true that on average black Africans in Africa score extremely low on scholastic/intelligence tests because they grow up with much less educational and other modern cultural resources (as Flynn would agree), then they should perform "extremely well" (by comparison) in those "g-loaded" cognitive contests that do not require too much of such quality cultural exposure (as Jensen would agree). (Chisala 2021)


Chanda argues that raw cognitive ability is better measured in Africa by a Scrabble championship than by an IQ test, since most Africans lack "access to well-trained teachers, big libraries, computers or even TVs" (Chisala 2021). Africans are good at Scrabble:


Nigeria happens to be the world's top performing nation in English Scrabble, while francophone African countries are also the most dominant in French Scrabble, despite the fact that the top players in Western countries are super-high-IQ nerds with visibly exceptional mathematical talents (Chisala 2021)


Correlation isn't causation. Is a high IQ needed to do well at Scrabble? Not according to this study:


Forty tournament-rated SCRABBLE players (20 elite, 20 average) and 40 unrated novice players completed a battery of domain-representative laboratory tasks and standardized verbal ability tests. The analyses revealed that elite- and average-level rated players only significantly differed from each other on tasks representative of SCRABBLE performance. Furthermore, domain-relevant practice mediated the effects of SCRABBLE tournament ratings on representative task performance, suggesting that SCRABBLE players can acquire some of the knowledge necessary for success at the highest levels of competition by engaging in activities deliberately designed to maximize adaptation to SCRABBLE-specific task constraints. (Tuffiash, Roring, and Ericsson 2007)


Success at Scrabble seems to be due largely to practice and is thus a poor measure of raw cognitive ability.


A curious detail: Nigeria's top performers come overwhelmingly from one part of the country: the Niger Delta, which is home to the Igbo and related tribes. Since the peoples of the Niger Delta used to dominate trade between the coast and the interior, and since trade selects for cognitive ability, mean IQ should be higher in those populations that have long practiced it, like the Igbo (Frost 2015).


Young Nigerians in the UK - Academic achievement on the GCSE


Although many African immigrants do poorly in British schools, some actually do well. A study of six secondary schools in inner London found that results on the General Certificate of Secondary Education (GCSE) were higher for African students who spoke Igbo, Yoruba, Luganda, and Ga than for White British students who spoke only English (Demie 2013, p. 9). Chanda sees the GCSE as a proxy for IQ and argues that IQ differences between African immigrants and White British must be highly malleable:


Africans speaking Luganda and Krio did better than the Chinese students in 2011. The igbo were even more impressive given their much bigger numbers (and their consistently high performance over the years, gaining a 100 percent pass rate in 2009!). The superior Igbo achievement on GCSEs is not new and has been noted in studies that came before the recent media discovery of African performance. A 2007 report on "case study" model schools in Lambeth also included a rare disclosure of specified Igbo performance (recorded as Ibo in the table below) and it confirms that Igbos have been performing exceptionally well for a long time (5 + A*-C GCSEs); in fact, it is difficult to find a time when they ever performed below British whites. (Chanda 2015)


Igbo students stood out as high achievers on the GCSE, as did Yoruba students to a lesser extent. In both groups, however, the mean results were highly variable from one year to the next:


2009: Igbo - 100%, Yoruba - 39%

2010: Igbo - 80%, Yoruba - 68%

2011: Igbo - 76%, Yoruba - 75% 

(Demie 2013, p. 9)


Chanda attributes this variability to statistical noise caused by small sample size. If so, there should be an inverse correlation between sample size and variability. GCSE scores should be more variable for smaller groups than for larger ones. Yet the reverse seems to be true for the years 2009 to 2011:


Yoruba: 90 students / gain of 36 percentage points

Somali: 53 students / gain of 13 percentage points

Twi-Fante: 37 students / loss of 3 percentage points

Igbo: 16 students / loss of 24 percentage points

Krio: 12 students / gain of 4 percentage points

Tigrinya: 12 students / loss of 8 percentage points

Lingala: 12 students / loss of 5 percentage points

Ga: 8 students / gain of 9 percentage points

Swahili: 8 students / gain of 10 percentage points 

(Demie 2013, pp. 7, 9)


The two largest gains were made by the two largest groups: the Yoruba and the Somali. If the differences between 2009 and 2011 are statistical noise, why are the largest ones associated with the largest groups? Shouldn't we see the reverse? Shouldn't the smallest groups show the most variability?


Something seems to be causing those impressive GCSE gains. Since the students are not the same from one year to the next, and since the gains differ considerably from one ethnic community to another, the "something" must be the community itself. Over time, the Yoruba community became better at assisting its students, and this kind of assistance was available only in larger communities like the Yoruba.


The most obvious forms of assistance are tutoring and coaching. Such assistance is mentioned by parents in interviews for the above study:


Parent A: Father of daughter in Year 9. Generally supportive of the school which was not his first choice but is supplementing his daughter's education with a home tutor. He also calls on his extended family, his oldest son who is a graduate is also expected to help. (Demie 2013, p. 14)


Although tutoring and coaching are perfectly legitimate, they invalidate the GCSE as a means to measure IQ, particularly its genetic component. If some students get tutored and others do not, there will be more environmental variance in IQ, and the exam results will say less about the genetic potential for cognitive ability. Therefore, GCSE results tell us what we already know: if you get tutored and coached before an exam, you'll do better.


Are tutoring and coaching the only forms of community assistance? There is another one: impersonation. In other words, the parents hire a smart student from their community to take the exam in their child's place. This strategy is feasible only if the community has enough individuals who are (1) intelligent and (2) similar in age and appearance to the student in question. Such individuals are lacking in a small community, as are the middlemen who can refer an anxious parent to a suitable source of assistance.


How common is this strategy? Adebayo (2013) studied cheating behavior among Nigerian university students and British university students. He found that impersonation services were used or provided by 20% of the former and 1% of the latter. In general, cheating took non-collaborative forms among British students and collaborative forms among Nigerian students:


These include behaviours like writing somebody's coursework, colluding with others to communicate answers to one another, over marking one another's course work etc. This is quite different from plagiarism and non-collaborative cheating characteristic of the British sample reported by Newstead et al (1996). Reasons for these differences may be attributable to differences in population, differences in cultural ethnic, differences in emphasis placed on examination as part of educational assessment (Adebayo 2013, p. 146)


Adebayo (2013, p. 148) found high rates of collaborative cheating among Nigerian students:


Permitting own coursework to be copied - 72.6%

Copying another student's coursework with consent - 47.3%

Collaborative generous marking of coursework - 64.6%

Submitting joint work as an individual's - 49.3%

Doing another student's coursework for them - 77.3%

Collusion with another student to communicate answers - 83%


We live in a world that has low-trust and high-trust societies. In a high-trust society, like the UK, cheating is considered shameful and disreputable, regardless of whom you cheat. In a low-trust society, like Nigeria, cheating is wrong only when you do it to friends and relatives.


What happens when individuals from a low-trust society migrate to a high-trust one? If they come in sufficient numbers, their opportunities for collaborative cheating are greatly increased. Imagine you're supervising an exam in an English school, and you suspect an African student is filling in for another. He shows you his school card and another piece of ID. Both are correct. So what do you do now? Do you really want to make a fuss and risk being accused of racial profiling? No you don't.


Future research


The GCSE study by Demie (2013) leaves much to be desired. It does not provides the number of students who had to retake that exam (which must be a large number); nor does it provide a breakdown of the number of students taking it per year.


In any case, the GCSE is a poor substitute for an IQ test. Even if we exclude cheating, the results are distorted by legitimate activities like tutoring and coaching. The latter are more available to some students than to others. Consequently, GCSE results tell us nothing about differences in raw cognitive ability, either between individuals or between communities.


Chanda promises to write an article that will rule out cheating as an explanation for Nigerian success on the GCSE. Again, the issue isn't just cheating. It's any assistance that goes to some students and not to others. If you want to measure raw cognitive ability, you need a level playing field. In particular, you need a test that does not offer high achievers the lure of personal gain, which may push test-takers to do well by hook or by crook. In the UK, an African with good GCSE results has access to a wide range of good-paying jobs, in large part because of "diversity quotas" of one sort or another.


This motive comes out in interviews with the parents of African students:


● 'Without an education you cannot earn a decent salary, without qualifications you cannot get a good job. The best thing is to push your children as hard as you can.'

● 'Being a Black woman if you don't have education in this country, what job will you have to do, clean people's toilets?'  (Demie 2013, p. 13)


This subject should definitely be a research priority. We need IQ data on Nigerians, and not inadequate substitutes like GCSE scores. We also need data on alleles associated with educational attainment (i.e., polygenic scores). Furthermore, we need data on each of Nigeria's ethnic groups, particularly the Igbo. It's hard to fake intelligence in the real world, and the Igbo have a long history of doing better at business and other endeavors. Unfortunately, intelligent people are also better at cheating, so there is some confounding between real intelligence and the fake kind.




Adebayo, S.O. (2011). Common Cheating Behaviour among Nigerian University Students: A Case Study of University of Ado-Ekiti, Nigeria. World Journal of Education 1(1): 144-149.


Chisala, C. (2015). The IQ Gap Is No Longer a Black and White Issue. The Unz Review, June 25


Chisala, C. (2020). Nigerians, Jews and Scrabble: An Update on the IQ Debate. The Unz Review, February 27


Demie, F. (2013). Raising Achievement of Black African Pupils. Good Practice in Schools. London: Lambeth Research and Statistics Unit, Lambeth Council.


Frost, P. (2015). The Jews of West Africa. The Unz Review, July 4


Tuffiash, M., R.W. Roring, and K.A. Ericsson. (2007). Expert performance in SCRABBLE: Implications for the study of the structure and acquisition of complex skills. Journal of Experimental Psychology: Applied, 13(3), 124-134.


Sean said...

"It's hard to fake intelligence in the real world, and the Igbo have a long history of doing better at business and other endeavors."

Does not seem to have produced much in the way of actual technical achievements, this Nigerian DNA currently being injected into the gene pool of Britain. A century ago the British Prime Minister Ramsay MacDonald was part Indian and the Admiral John Fisher part Malay, but what is going to happen when the cities are substantially African is not going to be integration because they will be in competition with Indians and Muslims in the professions. The sort of substantially Irish lumpenproletariat that inhabit English cities will be the ones mating with blacks.

Peter Frost said...


Keep in mind that the Igbo are a minority within a minority. I also believe that intelligence isn't everything. In particular, I believe that the creation of complex societies required many different cognitive and behavioral changes, cf. the studies by Gregory Clark.

I'm wary of predicting the future. However, if current trends continue, we will probably see a process of demographic replacement. The cities will no longer be British in any meaningful sense.

Bruce said...


Is there evidence of selection bias (I hope that’s the correct phrase)? By that I mean that the immigrant Nigerians (already disproportionately composed of the high intelligence Igbo) are skewed towards high performers. I assume average and below average Africans tend not to (legally) immigrate.