It is a measure of the quality of British life that one of its longest running TV programs is “University Challenge”, a quiz show for university students. Yes, it has always been a minority interest, but it is a showcase of talent, an astounding example of what bright young people can get to know in roughly 25 years.
My introduction to this phenomenon was at the University of Keele, which won the contest in 1968, the year of my graduation. I knew team member Pam Maddison (Groves), who studied Psychology in the year below me, and once chatted with me about the estimated number of objects in the universe, a concept I found mind-blowing. I think she added that she had checked her calculations with those of her boyfriend, and found them a reasonable match. It is good to meet bright people.
Unusually, University Challenge has stuck to the same format since inception in 1962, and that means that the long series of results is broadly comparable in the best sense of being an open competition following the same rules. Each contesting university achieves a score against a competitor, and at the final the winner beats the runner up, and their scores show the winning margin. In more detail, some contestants are better than others, and answer more questions, which gets them fan club status, like Eric Monkman (pictured above). Universities field their best candidates, having selected them in qualifying rounds, and then train up a small group before the chosen 4 go forward to compete. The winning team in each round is the one with the most correct answers and fewest incorrect interruptions. The subject matter is extremely broad, but the results of each competition are on ratio scales with true zeros and equal intervals.
Naturally, only the brighter people get to university, and presumably only the brighter of those get onto the university team. Many of the contestants have gone on to notable achievements in public life. The competition tests knowledge, plus the capacity to quickly judge from the question what answer is being looked for, and whether it is worth jumping in with the likely answer in order to obtain bonus points. In the psychometric jargon this is mostly about crystallized intelligence rather than fluid, on-the-spot problem-solving, intelligence, of the sort involved in mental calculations as showcased in another TV program, Countdown.
So, although University Challenge may not be the hardest test of raw intellectual power, it certainly demands very high ability. Adrian Furnham and colleagues found that IQ was the best predictor of general knowledge, but that Openness to Experience, a weak proxy for intelligence included in Five Factor personality assessments, made an additional contribution.
Cognitive ability, learning approaches and personality correlates of general knowledge
Adrian Furnham , Viren Swami , Adriane Arteche & Tomas Chamorro‐Premuzic Pages 427-437 | Received 05 Jul 2007, Accepted 03 Oct 2007, Published online: 20 May 2008
So, this is a showcase of talent, and it is great that it has built up a loyal following, and entered into British popular culture: “Fingers on buzzers” “Your starter for 10”. That is real fame, and I hope those phrases last. So, to use another phrase, though not from this TV program “What’s not to like?”
Michael Hogan, writing for “The Telegraph” (on the political Right) says:
Gender balance needs to be tackled next series
The all-male line-up for this final has sparked a sexism debate over the past week. The statistics are indeed pretty damning. One-third of this year’s teams had no women, only 22 per cent of the contestants this series were female and just five per cent of finalists over the last five years have been women. Equality quotas are a tricky topic but perhaps the team selection process within universities needs to be looked at and guidelines issued. Eight males on-screen in this showcase final – 10 if you include Paxman and Hawking – simply just doesn’t send out the right message.
Eve Livingston writing for “The Guardian” (on the political Left) says:
The year is 2017 and at 8pm on Monday 10 April, televisions across the country switch on to the BBC: four men from Oxford face four men from Cambridge in a combative race to prove their superior intelligence. Verbose questions and bellowed answers are punctuated only by sneering quips from a white male Cambridge alumnus. No, it’s not a parliamentary debate, but the final of University Challenge, a stalwart of middle-class British culture since the early 1960s.
The all-male contest ended a series in which just 22% of competitors were women, a fact that hasn’t gone unnoticed by viewers and campaigners alike.
University Challenge is, of course, hardly at the pinnacle of gender inequality issues facing women every day. And the show’s problems with representation don’t stop at gender – notably the under-representation of black students and the elitism inherent in questions about classical composers, Greek mythology and Renaissance literature, which see various iterations of Oxbridge colleges dominate year on year.
a conversation about representation is worthwhile. The BBC has previously said that institutions are responsible for selecting their own teams, effectively laying responsibility at the door of universities (and some do voluntarily put in place quotas), but there is nothing to stop the broadcaster from issuing guidelines or conditions for entry. In the meantime, an acknowledgement of the problem from both parties and a meaningful commitment to tackling it would be a good starter for 10.
It would appear we have a consensus that something has gone wrong, and both ends of the political spectrum are asking for quotas. I have not checked these figures, but the final winning teams since inception number 184 contestants, of whom only 16 were women, so their representation is roughly 9%.
I am not writing for a national newspaper, but I take a more measured approach than to ask for quotas. What do we know about general knowledge and sex differences outside this particular TV format?
Sex differences in general knowledge, semantic memory and reasoning ability
Richard Lynn and Paul Irwing, British Journal of Psychology (2002), 93, 545–556.
This paper has the three objectives of attempting to replicate a previous study in which it was found that males have substantially greater general knowledge in most fields or domains than females, and of determining how far sex differences in general knowledge are a function of differences in either Gf (fluid intelligence), or experience. The results confirmed the previous study to the effect that males have higher means in a general knowledge factor of approximately .50d (half a standard deviation). It was found further that there was no significant sex difference in Gf measured by Baddeley’s Grammatical Reasoning Test, and only a low correlation between general knowledge and Gf. Analysis of covariance showed that differential experience as indicated by ‘A’-level points and socio-economic status had only a marginal impact on the observed sex difference. The results are interpreted as showing that sex differences in general knowledge cannot be explained as a function of differences in either Gf or experience. It is proposed further that general knowledge should be regarded as a new second-order factor and designated as semantic memory.
Lynn and Irwing argue that men have always been better at the Information (general knowledge) subtest of the Wechsler test, an important finding because the of the care taken over the representativeness of the standardization sample. Furthermore, boys are better at girls on wider general knowledge in 26 European countries. How general is general knowledge? The authors search for the underlying domains, in order to be sure that their knowledge test is not sex-biased.
In the work of the present authors, tests were constructed of 19 domains of general knowledge and factored to produce six first-order factors. These consisted of Physical Health and Recreation (games, biology and sport); Current Affairs (politics, history, geography, exploration, finance); Family (cookery and medicine); Science (general science and history of science); Fashion (clothes fashion, film, pop music); and Arts (classical music, visual art, jazz and literature). These first-order factors yielded a general factor. The sex differences were that males significantly outperformed females on all the domains of the Physical Health and Recreation factor (biology, games and sport); on all the domains of the Current Affairs factor (politics, finance, history, exploration, geography); and on all the domains of the Science factor (general science, history of science). Females significantly outperformed males on both domains of the Family factor (medicine and cookery). There were no statistically significant differences on the domains of the Fashion factor (clothes fashion, popular music, film). On the Arts factor, males significantly outperformed females on the domains of literature and jazz, while there were no statistically significant differences on visual art or classical music. As in the Ackerman studies, males outperformed females on the majority of the tests. The sex difference on the entire battery of tests was .51 d favouring males. This difference is even greater than the differences normally found on the Weschler standardization samples. It is so large that it requires replication and this is the first objective of the study reported here.
Their sample is sizeable, and seems representative:
The sample comprised 1047 undergraduate students (594 women and 453 men) from the Faculties of Science; Informatics; Engineering, Arts; and Health, Social Sciences and Education, at the University of Ulster, who ranged in age from 17 to 48 years (M= 20.5, SD = 3.3). The sex composition of the sample was representative of the student body.
While there was a significant male advantage with respect to total ‘A’-level points attained (female mean = 14.7, male mean = 16.9, t (721) = 4.7, p< .001), men and women did not differ significantly with respect to age (female mean = 20.4, male mean = 20.6, t (1037) = 0.88, p> .05), scores on Baddeley’s Grammatical Reasoning Test (female mean = 30.7, male mean = 29.7, t (1046) = 1.2, p> .05) or socio-economic status as indicated by father’s education (x2(3) = 0.9, p> .05) and occupation (x2(2) = 1.6, p > .05; see Table 1). As indicated by these data, the University of Ulster recruits a very high proportion of its students from groups of lower socio-economic status compared with most UK universities.
Discussing their results, they say:
It was found that the male advantage on the general factor of general knowledge is .48 d and is virtually identical to our previous result. The present study used a shorter form of the general knowledge test than that used in our previous study. When the sample used in the previous study is scored for the shorter form of the test, the sex difference is .46d. This again is virtually identical to the difference of .48 d obtained on the same shorter form of the test in the present study. Our results are also similar to those obtained in the USA by Bowenet al. (2000). Our first conclusion is therefore that the magnitude of the sex difference on the general factor of general knowledge of approximately .50 d is a robust and replicable result.
They make further points about general knowledge and how it fits in the hierarchy of abilities, showing that fluid intelligence measures are only weakly correlated with it, but the main point is that the observed male advantage in University Challenge is not an artefact of selection for a TV program, but an established aspect of sex differences in knowledge. Since men are better at general knowledge, and are usually more variable in ability (larger standard deviations) than women it would make sense that there would be fewer women selected for local university team membership, and progressively far fewer in winning teams. As you push out towards higher levels of general knowledge there are about 10 very knowledgeable men for every equally knowledgeable woman.
This possibility either does not occur to journalists because they do not know the basic facts, or is suspected but has been dismissed by them as something not to be expressed in mixed company.
Disclaimer: I would not have been able to get onto the University of Keele team in any year. The best I have ever done while watching the program is 8 answers correct. That was without counting mistaken interruptions.