Ryan Faulk, like most IQ-ists, believes that the correlation with job performance and IQ somehow is evidence for its validity. He further believes that because self- and peer-ratings correlate with one’s IQ scores that that is further evidence for IQ’s validity.
Well too bad for Faulk, correlations with other tests and other IQ tests lead to circular assumptions. The first problem, as I’ve covered before, is that there is no agreed-upon model or description of IQ/intelligence/’g’ and so therefore we cannot reliably and truthfully state that differences in ‘g’ this supposed ‘mental power’ this ‘strength’ is what causes differences in test scores. Unfortunately for Ryan Faulk and other IQ-ists, again, coming back to our good old friend test construction, it’s no wonder that IQ tests correlate around .5—or so is claimed—with job performance, however IQ test scores correlate at around .5 with school achievement, which is caused by some items containing knowledge that has been learned in school, such as “In what continent is Egypt?” and Who wrote Hamlet?” and “What is the boiling point of water?” As Ken Richardson writes in his 2017 book Genes, Brains, and Human Potential: The Science and Ideology of Intelligence (pg 85):
So it should come as no surprise that performance on them [IQ tests] is associated with school performance. As Robert L. Thorndike and Elizabeth P. Hagen explained in their leading textbook, Educational and Psychological Measurement, “From the very way in which the tests were assembled [such correlation] could hardly be otherwise.”
So, obviously, neither of the two tests determine independently that they measure intelligence, this so-called innate power, and because they’re different versions of the same test there is a moderate correlation between them. This goes back to item analysis and test construction. Is it any wonder, then, why correlations with IQ and achievement increase with age? It’s built into the test! And while Faulk does cite high correlations from one of Schmidt and Hunter’s meta-analyses on the subject, what he doesn’t tell you is that one review found a correlation of .66 between teacher’s assessment and future achievement of their students later in life (higher than the correlation with job performance and IQ) (Hoge and Coladarci, 1989.) They write (pg 303): “The median correlation, 0.66, suggests a moderate to strong correspondence between teacher judgments and student achievement.” This is just like what I quoted the other day in my response to Grey Enlightenment where I quoted Layzer (1972) who wrote:
Admirers of IQ tests usually lay great stress on their predictive power. They marvel that a one-hour test administered to a child at the age of eight can predict with considerable interest whether he will finish college. But as Burt and colleagues have clearly demonstrated, teachers subjective assessments afford even more reliable predictors. This is almost a truism.
So the correlation of .5 between occupation level and IQ is self-fulfilling, which are not independent measures. In regard to the IQ and job performance correlation, which I’ve discussed in the past, studies in the 70s showed much lower correlations, between .2 and .3, which Jensen points out in The g Factor.
The problem with the so-called validity studies carried out by Schmidt and Hunter, as cited by Ryan Faulk, is that they included numerous other tests that were not IQ tests in their analysis like memory tests, reading tests, the SAT, university admission tests, employment selection tests, and a variety of armed forces tests. “Just calling these “general ability tests,” as Schmidt and Hunter do, is like reducing a diversity of serum counts to a “general. blood test” (Richardson, 2017: 87). Of course the problem with using vastly different tests is that they tap into different abilities and sources of individual differences. The correlation between SAT scores and high school grades is .28 whereas the correlation between both the SAT and high school grades and IQ is about .2. So it’s clearly not testing the same “general ability” that’s being tested.
Furthermore, regarding job performance, it’s based on one measure: supervisor ratings. These ratings are highly subjective and extremely biased with age and halo effects seen with height and facial attractiveness being seen to sway judgments on how well one works. Measures of job performance are unreliable—especially from supervisors—due to the assumptions and biases that go into the measure.
Do IQ tests test neural processes? Not really. One of the most-studied variables is reaction time. The quicker they react to a stimulus, supposedly, the higher their IQ is in average as they are quicker to process information, the story goes. Detterman (1987) notes that other factors other than ‘processing speed’ can explain differences in reaction time, including but not limited to, stress, understanding instructions, motivation to do said task, attention, arousal, sensory acuity, confidence, etc. Khodadadi et al (2014) even write “The relationship between reaction time and IQ is too complicated and reveal a significant correlation depends on various variables (e.g. methodology, data analysis, instrument etc.).” Complex cognition in real life is also completely different than the simple questions asked in the Raven (Richardson and Norgate, 2014).
It is easy to look at the puzzles that make up IQ tests and be convinced that they really do test brain power. But then we ignore the brain power thst nearly everyone displays in their everyday lives. Some psychologists have noticed thst people who stumble over formal tests of cognitive can bangle highly complex problems in their real lives all the time. As Michael Eysenck put it in his well-known book Psychology, “There is an apparent contradiction between our ability to deal effectively with out everyday environment and our failure to perform well on many laboratory reasoning tasks.” We can say the same about IQ tests.
Real-life problems combine many more variables that change over time and interact. It seems that the ability to do pretentious problems in a pencil-and-paper (or computer) format, like IQ test items, is itself a learned, if not-so-complex skill. (Richardson, 2017: 95-96)
Finally, Faulk cites studies showing that how intelligent people and their peers rates themselves and others predicted how well they did on IQ tests. This isn’t surprising. Since they correlate with academic achievement at .5 then if one is good academically then they’d have a high test score more often than not. That friends rate friends high and they end up matching scores is no surprise either as people generally group together with other people like themselves and so therefore will have similar achievements. That is not evidence for test validity though!! See Richardson and Norgate (2015) “In scientific method, generally, we accept external, observable differences as a valid measure of an unseen function when we can mechanistically relate differences in one to diffences in the other …” So even Faulk’s attempt to ‘validate’ IQ tests using peer- and self-ratings of ‘intelligence’ (whatever that is) falls on its face since its not a true measure of validity. It’s not construct validity. (EDIT: Psychological constructs are validated ‘by testing whether they relate to measures of other constructs as specified by theory‘ (Strauss and Smith, 2009). This doesn’t exist for IQ therefore IQ isn’t construct valid.)
In sum, Faulk’s article leaves a ton to be desired and doesn’t outright prove that there is validity to IQ tests because, as I’ve shown in the past, validity for IQ is nonexistent, though some have tried (using correlations with job performance as evidence) but Richardson and Norgate (2015) take down those claims and show that the correlation is between .2 and .3, not the .5+ cited by Hunter and Schmidt in their ‘validation studies’. The criteria laid out by Faulk does not prove that there is true construct validity to IQ tests and due to test construction, we see these correlations with educational achievement.