Some readers may have fun reading this. I have not edited anything. Due to the conversion some extra lines were added, but his emails genuinely contains lots of oddly placed line breaks. The odd use of e.g. bold is his (Jayl is a male name, apparently).
The topic of the email is “University of Toronto, Canada”.
| Jayl Feynman <feynmantoronto@outlook.com> |
5. oktober 2015 kl. 23.01 |
|
Til: “emil@emilkirkegaard.dk” <emil@emilkirkegaard.dk>
|
I am a cognitive neuroscience researcher from Toronto and wanted to make comment if you don’t mind on one of your studies titled “ Sex differences in g and chronometric tests”
Study – file:///C:/Users/user/Downloads/article2.pdf
“Males average 100g more brain than women, and brain size is known to correlate with general intelligence (g), leading to the possibility that men average somewhat higher in g than women”
Intelligence is a reflection of the efficacy of the networks responsible for cognition and not merely the absolute size of the human brain. Hence it’s quality over quantity. The best model for predicting the neuroscience of intelligence is the P-Fit theory which you could read about it here:
http://www.livescience.com/1863-theory-intelligence-works.html
The P-Fit theory also contradicts the claim that G factor does not change as Parietal-frontal integration changes through experience and connectivity.
“It has also been shown that brain size correlates with IQ with a strength of about .38“
Correlation is not causation. Even Einstein’s brain was smaller than average:
1. IQ tests such as Raven progressive matrices are very much flawed because almost all of the questions are spatial in nature (picture completion,block design etc) and do not really touch on the verbal part of intelligence. To date, I have never seen an IQ test that measure verbal comprehension or fluency Measuring spatial ability also cannot be predicative in measuring ability in comprehending literature, novels, books and just about everyday language use.
2. Psychologists do not distinguish between different forms of information when measuring IQ. Different forms of information such as spatial and verbal information uses different areas of the brain for example pictures shapes and even arithmetic or digits use an area of the brain known as the Parietal lobe. Numeracy and digits which Psychologists think are verbal actually happens to be spatial since the brain represents them as quantity of space through mental imagery. Verbal information on the other hand such as words, sentences, language and meaning are processed by two areas of the brain known as Brocha’s and Wernicke’s area. Therefore IQ does not measure verbal intelligence especially in processing, reasoning and understanding the use of language.
3. Sample restriction and recruitment bias. The methods in which members of the each group enter the study sample is not in the control of the investigator in charge of the study. For example, in college or school studies it could easily be that both average and less average females enter the study while higher level males enter the study. This would obviously obscure mean differences.
4. Variability hypothesis and bias. This hypothesis holds that males exhibit greater variation than females in many cognitive ability domains, which may explain their overrepresentation in the tails of ability distributions and creates the appearance of mean differences in incomplete or selected samples. The male variability is obviously higher in the right tail of college samples therefore college sex differences in IQ is inaccurate in measuring IQs of the general population.
5. Intelligence is an emergent property of anatomically distinct networks of the brain each of which has it’s own distinct capacity. Therefore to measure general intelligence one would have to measure the distinct capacity of each brain network on which a singular IQ testing cannot do. You would need multiple separate tests to measure the capacity of each network.
|
|
| Emil Ole William Kirkegaard <the.dfx@gmail.com> |
6. oktober 2015 kl. 02.36 |
|
Til: Jayl Feynman <feynmantoronto@outlook.com>
|
Jayl,
I googled your name yet was unable to find anything about your academic standing. LinkedIn is the only result and states that you are “Entrapreneur at MySelf (as an independent consultant)”. There is no mention of your name on the university website and I could locate no publications in your name on Google Scholar. The only things that come up are the publications of a physicist doing work on e.g. solar physics. Your email comes from a non-university address whereas academics usually use their institutional email. They also usually have a signature, whereas you have none.
So, taken all together, the evidence seems to show that you are not who you say you are. I’ll make the assumption that you are a random internet person who is uncomfortable with sex differences in cognitive ability. After all, many people are.
Anyway, I will reply, just for fun. :)
—
The paper you ‘link to’ (actually you gave the file’s location on your own computer) is a submission that was never published. You can read the submission thread to see why. http://openpsych.net/forum/showthread.php?tid=9
Intelligence is a reflection of the efficacy of the networks responsible for cognition and not merely the absolute size of the human brain. Hence it’s quality over quantity. The best model for predicting the neuroscience of intelligence is the P-Fit theory which you could read about it here:
Correlation is not causation. Even Einstein’s brain was smaller than average:
Why do you presume I don’t know what P-FIT is? If so, linking me to unreliable secondary sources is even more strange. Good rule of thumb: if you want to introduce a researcher to something, cite the primary literature. In this case:
Jung, R. E., & Haier, R. J. (2007). The Parieto-Frontal Integration Theory (P-FIT) of intelligence: converging neuroimaging evidence. Behavioral and Brain Sciences, 30(02), 135-154.
There are also some newer relevant work, e.g. by Roberto Colom.
In any case, I don’t think anyone made the claim that size explains all variation in cognitive ability, only that it explains some of it. Given the evidence, it would be extremely surprising if it did not. The very thing that makes humans unique is their high cognitive ability which increased over recent evolutionary time along with the brain size (i.e. we see it by increasing skull sizes; relative to body size).
If brain size was a non-causal correlate, one has the odd job of explaining why it was selected for by evolution just over the same time span. Note that brain tissue is extremely metabolically expensive. The brain accounts for about 20% of the rest metabolic rate yet takes up only about 2% by weight. Increased brain size causes substantial problems with childbirth which kills off a large number of women. A non-causal theory is that evolution still selected for larger brains despite all these costs. Presumably, this is why pretty much no one serious subscribes to that idea.
As for Einstein’s brain. You are citing a single case which cannot disprove an imperfect correlation, so the case is a non-starter. In this case, the case has an obvious explanation: the brain shrinks as we age (matching the decline in absolute scale cognitive ability).
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2596698/
—
The P-Fit theory also contradicts the claim that G factor does not change as Parietal-frontal integration changes through experience and connectivity.
No one claims a stability of 1 for GCA, hence change in brain structure is clearly consistent with relatively stable GCA. Also note that our measures of GCA measure relative standing (they are deviation scores), not absolute ability. So in fact, no change in GCA is consistent with change in the underlying brain structure if all members of the cohort have the exactly same brain changes (this isn’t true, but I’m just pointing it that the argument to be valid, needs this unstated premise).
—
This looks to be the stuff you really wanted to say. IQ criticism.
1. IQ tests such as Raven progressive matrices are very much flawed because almost all of the questions are spatial in nature (picture completion,block design etc) and do not really touch on the verbal part of intelligence. To date, I have never seen an IQ test that measure verbal comprehension or fluency Measuring spatial ability also cannot be predicative in measuring ability in comprehending literature, novels, books and just about everyday language use.
2. Psychologists do not distinguish between different forms of information when measuring IQ. Different forms of information such as spatial and verbal information uses different areas of the brain for example pictures shapes and even arithmetic or digits use an area of the brain known as the Parietal lobe. Numeracy and digits which Psychologists think are verbal actually happens to be spatial since the brain represents them as quantity of space through mental imagery. Verbal information on the other hand such as words, sentences, language and meaning are processed by two areas of the brain known as Brocha’s and Wernicke’s area. Therefore IQ does not measure verbal intelligence especially in processing, reasoning and understanding the use of language.
In fact, RPM does not have picture completion or block design. These are both subtests of the WAIS.
You say you have never seen an IQ test measuring these, which means that you have not looked hard because WAIS, probably the most widely used test of all, has a an entire subscale called verbal comprehension consisting of 3-4 tests, depending on which version of WAIS.
http://images.pearsonclinical.com/images/assets/WAIS-IV/WAISIV2_6_08.pdf
The last claim about lack of cross-area predictive validity is widely known to be false. Indeed, it has been known for decades, even going back to Spearman’s time in the 1930s. This is what is called indifference of the indicator, it doesn’t matter that much just which mental test is used to assess GCA, as long as the g-loading is strong, the predictive validity will be similarly strong. A great general resource is still:
—
3. Sample restriction and recruitment bias. The methods in which members of the each group enter the study sample is not in the control of the investigator in charge of the study. For example, in college or school studies it could easily be that both average and less average females enter the study while higher level males enter the study. This would obviously obscure mean differences.
4. Variability hypothesis and bias. This hypothesis holds that males exhibit greater variation than females in many cognitive ability domains, which may explain their overrepresentation in the tails of ability distributions and creates the appearance of mean differences in incomplete or selected samples. The male variability is obviously higher in the right tail of college samples therefore college sex differences in IQ is inaccurate in measuring IQs of the general population.
Sometimes the researcher do get to decide, especially when they collect their own data. Other times, they use available datasets which may have recruitment bias. The fact that you cite research written by top researchers (Earl Hunt, Ian Deary) in the field discussing this means that they are aware of the problem.
—
5. Intelligence is an emergent property of anatomically distinct networks of the brain each of which has it’s own distinct capacity. Therefore to measure general intelligence one would have to measure the distinct capacity of each brain network on which a singular IQ testing cannot do. You would need multiple separate tests to measure the capacity of each network.
You seem unaware of the g factor and seem to posit something like the Thompson’s sampling theory. The networks are not independent and attempts to make tests that do note correlate have all failed despite decades of attempts. Maybe read:
In any case, WAIS does try to measure a diverse set of cognitive abilities. This is a good idea because it results in a better measurement of GCA, which is what is responsible for the validity of the tests. See review at:
https://emilkirkegaard.dk/en/?p=4581
-Emil
|
|
| Jayl Feynman <feynmantoronto@outlook.com> |
6. oktober 2015 kl. 04.20 |
|
Til: Emil Ole William Kirkegaard <the.dfx@gmail.com>
|
“The very thing that makes humans unique is their high cognitive ability which increased over recent evolutionary time along with the brain size (i.e. we see it by increasing skull sizes; relative to body size)“
The prefrontal cortex is slightly larger relative to the rest of the brain in humans than in most other primates while also having larger volume of white matter to go alongside within it.
“If brain size was a non-causal correlate, one has the odd job of explaining why it was selected for by evolution just over the same time span”
Because bigger brains equals to more lateralization (asymmetry) of brain functions.
file:///C:/Users/user/Downloads/fnhum-08-00915.pdf
Lateralized brain allows dual attention to the tasks of feeding (right eye and left eye hemisphere) and vigilance for predators (left eye and right hemisphere). Hence it was most likely selected because dual attention was an advantage for human survival.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3973910/
|
|
| Jayl Feynman <feynmantoronto@outlook.com> |
6. oktober 2015 kl. 06.03 |
|
Til: Emil Ole William Kirkegaard <the.dfx@gmail.com>
|
|
|
| Jayl Feynman <feynmantoronto@outlook.com> |
8. oktober 2015 kl. 22.54 |
|
Til: Emil Ole William Kirkegaard <the.dfx@gmail.com>
|
So you didn’t respond huh? Well then I will just end with this, after I looking over different test samples I have come to the conclusion that IQ tests are nonsense. The questions are almost all spatial in nature while having a quantitative and visuo-spatial section is redundant as they both measure mental rotation and spatial visualization. IQ test should thus be re-named SIQ or Spatial intelligence quotient because that is precisely what it measures. School academics are probably better predictors for future success more than this pop quiz. So I guess psychology have been getting it wrong for the last 100 years, but what can you expect from a field of study that only looks for correlations right? Well you can find correlations in anything for example I correlate with Psychology with bigotry since it produce the most out of any field example Rushton, Lynn and probably you.
Chao :P
|
|
I guess he thinks e.g. these are spatial items:
(All from Jensen’s Bias in Mental Testing).
The criticism is particular odd because I already told him about the actual composition of the WAIS.
Notice the links to files on his own computer. The same novice computer mistake made by the feminists writing for the UN.
I was not able to find the source paper for the claim that brain sizes got smaller during the last 30kya period. Can someone find it? Did body size shrink as well? If so, then brain-to-body size ratio may have increased over the period, or stayed the same.
I haven’t heard of the brain size evolved due to lateralization hypothesis before, but the review article he linked to seems interesting: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3973910/
The real reason I didn’t reply was that I was busy traveling home from the USA to Denmark. Still, it doesn’t seem worth my time.