Life is an IQ test (and so are video games)

 

 

Silliness will never end, and clever silliness will never stop getting credulous coverage. In the UK we have to contend with a particular scientist claiming that video games and computer tablets are changing children’s brains, and for the worse. Of course, in a minor way Psychological Comments is changing your brain, but for the better, I hope. When you read some nonsense about IQ in the newspapers you will mutter to yourself: “Wasn’t there a blog somewhere that took this research apart?” thus revealing that in some recess of your cerebral cortex there must be a physical change corresponding to your engraved memory of my wisdom. Brains change to some extent according to how they are used. I doubt any of these changes are as great as those inflicted by normal ageing, or by abnormal drinking of alcohol, but there remains a fear that new technology will change us. The plough changed us. Printing changed us. So did the first representational cave painting, the later adoption of perspective in painting, cinema, television, and video games.

So, what can we say about video games, beyond them being a colossal waste of time, on a par with playing Bridge and visiting relatives?

Quiroga, Escorial, Román, Morillo, Jarabo, Privado, Hernández, Gallego and Colom. Can we reliably measure the general factor of intelligence (g) through commercial video games? Yes, we can!   doi:10.1016/j.intell.2015.08.004

One hundred and eighty eight university undergraduates took part in the study. They played twelve commercial video games under strict supervision in the laboratory and completed eleven intelligence tests. Several factor models were tested for answering the question of whether or not video games and intelligence tests do measure the same underlying high-order latent factor. The final model revealed a very high relationship between the high-order latent factors representing video game and intelligence performance (r = .93). General performance scores derived from video games and intelligence tests showed a correlation value of .963 (R²_adjusted). Therefore, performance on some video games captures a latent factor common to the variance shared by cognitive performance assessed by standard ability tests.

remarkable relationships are only achieved when video games comprise moderate levels of complexity, display low consistency across practice sessions, and have no possibility of obtaining benefit from previously acquired skills. Importantly, some video games performance show a medium to high correlation with intelligence, even after extensive periods of practice, leading to the conclusion that performance on some video games cannot be automated (Quiroga et al., 2011).

https://drive.google.com/file/d/0B3c4TxciNeJZTHl4ZG9BZ0RoUXc/view?usp=sharing 

 

Unlike a reported 48% of the European population, I am not a gamer, but the games listed in this paper (categorized according to whether they mostly require analysis, visualisation, computing or memorization) seem deathly dull and educational. Years ago Douglas Adams kindly gave me an almost-ready version of his “Starship Titanic” to play with, which was fun, but these particular productions seem to reek of classroom dust. I thought computer games were about stealing fast cars, and then making a getaway. Perhaps that is romantic novels. Hard to keep up with these things.

Incidentally, the authors point out that: video games have been used for training (Granic, Lobel, & Engels, 2014; Green, Pouget, & Bavelier, 2010; Sitzmann, 2011) and for assessment purposes (Zygouris et al., 2014). As discussed by Granic et al. (2014) current video games are much more sophisticated than just a decade ago. Their review suggests that playing video games evoke widespread benefits in the cognitive, motivational, emotional, and social domains.

I particularly like the main finding, which is that good measures of intelligence can be derived from any mentally taxing everyday activity. In that sense we are achieving the goal of an anytime IQ test. Hernández-Orallo and Dowe (2010) Artificial Intelligence 174 1508–1539 Measuring universal intelligence: Towards an anytime intelligence test. It has 4 main features:

•The test should be able to measure the intelligence of any biological or artificial system that exists now or in the future.

•It should be able to evaluate both inept and brilliant systems as well as very slow to very fast systems.

•The test may be interrupted at any time, producing an approximation to the intelligence score, in such a way that the more time is left for the test, the better the assessment will be.

• It utilises the measurement of machine intelligence based on Kolmogorov complexity and universal distributions (a measure of the computational resources needed to specify an object, which were developed in the late 1990s (C-tests and compression-enhanced Turing tests).

It could be that the anytime test will come in the form of a protracted game, for a prize, which takes the contestant up the layers of difficulty, keeping a good record of multi-channel ability. That is the best case. The London tube shows that many passengers are playing versions of Patience, a dull game to distract themselves from a duller journey.

The researchers have shown that the medium is not the message. If computer games tax the intellect to any extent (and this selection has probably been chosen to do so) then they tap into the same common factor found in a range of more formal tests of intelligence. Yes, the subjects are the usual suspects: university students, but in this case the unrepresentative sample works against the hypothesis. The restriction of range will reduce g loadings, and brighter samples tend to have lover g scores anyway. Assuming that these students are above average in intelligence, then that will account for so many of the results departing from normal distributions, which has been corrected in the analysis, but remains a reality. In sum, finding a common factor is less likely in this selected population than it would be in the general population.

The results may seem predictable, but that is only because the positive manifold is so easy to replicate because……. it’s true.

Disclosure: I do not play video games. I played Canasta with my Granny as a child, but I doubt that counts for anything. I liked Pinball, but it was hardly an intellectual pursuit. My brain has probably been distorted by reading and talking to friends, but it is too late to repair that source of continuing damage.