IQ is a Valid Metric

IQ is able to predict job performance to some degree. Schmidt and Hunter (1990) examined the correlation between job performance and IQ Individual Differences in Output Variability as a Function of Job Complexity (gwern.net) . Methodologically speaking, they used coworker's assessments of work and tested it against metrics (such as IQ) to see correlation. IQ was the 2nd best metric, tying for second with a "Structured Interview", and beaten only by "Work Sample Tests" VA8gjB7.png (284×250) (imgur.com) . IQ beat things like "Job Experience," "Job Tryout," and even "Years of Education".

Strenze (2006) performed a meta-analysis of longitudinal studies (average group size being 97,083) that compared various life factors (e.g. IQ, Grades, SES index etc.) with education, occupation and income levels doi:10.1016/j.intell.2006.09.004 (emilkirkegaard.dk) . With education, occupation and income levels, IQ consistently produced the best correlation qIGdGqI.png (378×564) (imgur.com) . .

Clearly, IQ measures something that is valid, otherwise these decent correlations wouldn't be produced.

So, IQ (1) measures something valid (better than most other things), and (2) that valid thing it measures is probably intelligence.

LMAO - sure - let's let the people who "know what they're doing" run everything let's see how well that does. 

Supporters of IQ testing have been quick to point to correlations between IQ and job performance as evidence of test validity. A closer look at the data and results, however, suggests a rather murkier picture. Here we have acknowledged the methodological advances in meta-analyses from which such evidence has been drawn, while drawing attention to the problems surrounding them in this particular area. We conclude with a summary of the main points:

1. Much in developmental theory, and psychology in general, depends upon the validity of IQ tests.
   
2. In the absence of agreed construct validity this has weighed heavily on indirect validity using correlations with criterion outcomes among which job performance has a special status.
   
3. Hundreds of studies prior to the 1970s reported low and/or inconsistent correlations between IQ and job performance.
   
4. These correlations have been approximately doubled using corrections for supposed errors in primary results and combining them in meta-analyses. Such corrections have many strengths, theoretically, but are compromised in these cases by the often uncertain quality of the primary studies.
   
5. The corrections to sampling errors, measurement errors, and to range restriction have required making a number of assumptions that may not be valid and have created a number of persistently contentious issues.
   
6. The claim that the IQ-job performance correlation increases with job complexity is not born out in more recent studies.
   
7. A range of other—including noncognitive—factors could explain a correlation between IQ and job performance, and even constitute part or all of the enigmatic “general factor.”
   
8. There remains great uncertainty about the interpretation of IQ-job performance correlations and great caution needs to be exercised in using them as a basis for the validity of IQ tests and associated concepts.
   

As others have pointed out, statistical corrections are no magical compensation for weak data and that it is risky to reach conclusions about test validities from those currently available (Oswald & McCloy, 2003; Russell & Gilliland, 1995). The only solution is properly conducted primary studies, with larger representative samples, better measures, and so on. Until they are available, investigators should be extremely cautious about disseminating conclusions about IQ test validities, from correlations between IQ and job performance.

We don't even know if IQ tests can accurately predict job success much less if they're leaders. 

I think the results are repeatable enough to conclude that IQ measures some aspects of the human intellect but hardly encompasses what humans mean when we say intelligence.  We were just recently discussing Bobby Fisher's high IQ although that sad bastard was an unemployable schizophrenic who pissed himself to death because he wasn't smart enough to go to a doctor.

I do not disagree that IQ is a valid metric of something, just not intelligence.

I think the results are repeatable enough to conclude that IQ measures some aspects of the human intellect but hardly encompasses what humans mean when we say intelligence. 

We were just recently discussing Bobby Fisher's high IQ although that sad bastard was an unemployable schizophrenic who pissed himself to death because he wasn't smart enough to go to a doctor.  Was Bobby Fisher therefore intelligent?

I know a fair number of brilliant manic depressives but the people I know with a real genius for living an admirable life would probably score about 100 on that test.

I think IQ is over-rated and fairly slanted towards a 19th century European male model of success- lawyers and doctors and engineers.  What of empathy?  What of self-awareness? What of leadership and problem-solving?  What of wit and humor- clearly signs of intelligence  that no IQ test can capture.  What of instinct- which is often the smartest part of us though we don't understand how or why?   IQ might correlate with job performance because it favors those  intellects which dot their i's and cross their t's, but tidy thinkers are seldom inspired and the truly creative seldom score high in IQ.

We can recognize that dogs are smarter  than sheep but an IQ test can't tell us why.  Or why some dogs are smarter than others.  Clearly adults are smarter than children in most ways so intelligence changes and adapts as we grow and age.   Obviously, there's more to intelligence than can be expressed in letters and numbers and in fact those abstractions came along only recently, long after humans had distinguished themselves as more intelligent than any other species.

Overall, I'm IQ skeptical and my experience tells me (hey, what about experience?)  that applying numbers to people's capacity is often self-defeating or at least self-limiting.  Tell people that a 4 minute mile is the limit of human capability and for some reason everybody believes it until one day that record is broken and suddenly thousands more find that ability within themselves.

How would you define intelligence?

What was this "degree"?

Let's scrutinize these correlation coefficients as it concerns I.Q.:

Correlation with education level:

I.Q. : .49

Correlation with occupation level:

I.Q. : .41

Correlation with income level:

I.Q. : .22

And while the strength of correlation coefficients varies among researchers' metrics, these coefficients are weak to, at best, moderate.

I.Q. does not measure intelligence... It, at best, gauges classroom discipline.

I.Q. is as "valid" as a dance score, or rating of attractiveness.

Life outcomes are influenced by a plethora of social factors.

IQ is influenced by life outcomes and inherent predisposition.

Therefore research would suggest that life outcomes + predisposition = IQ = life outcomes.

Though a high IQ is no guarantee of social success.

Research and statistics are  indicative of  trends and expectancies....The normal expectancies relative to a particular social environment, as it were.

So basically, IQ is a valid metric of IQ expectancy, relative to social and physiological  factors that might or might not positively or negatively influence life outcomes.

So are we proving anything that we didn't already know?

I took the American SAT once. Irrelevant, I know, but I just want to say that it made absolutely no sense to me. I failed it while international schools enjoyed a huge jump because there was a sizeable Australian expats scoring high in the SAT. I'm guessing non-whites mingle with them to score well on it too. My point: tests are region-specific. They have no validity outside of their intended audience. The IQ test doesn't measure intelligence because if it did, the test would score accurately across different cultures. 

@oromagi

I do not disagree that IQ is a valid metric of something, just not intelligence.

Do you see any issue with the data I provided?

•  WwSHDHN.png (602×140) (imgur.com) 
  
• You claim that this is a snapshot from a Scientific American book but that particular data set has a much longer history.  I assume the article to which you refer is a reprint of Gottfredson's The General Intelligence Factor in which your table is reprinted.  Gottfredson is best known for leading the tepid scientific defense of Herrnstein and Murray's The Bell Curve (in which this table is also reprinted).
  
• Let's note that the Southern Poverty Law Center defines Gottfredson as a promoter of eugenicism, scientific racism, and white nationalism.
  
• Gottfredson has taken more than a quarter million dollars in grants from the White Supremacist Pioneer Fund to advance eugenics research.
  
• The table also appears in "The Bell Curve"
  
• Ultimately, this data comes from the  National Longitudinal Survey of Youth 1979 which began in 1979 with 12,686 men and women born in 1957-64 (ages 14–22).  Members were interviewed annually from '79 to '94 but this table was published in '94 so this data cover outcomes over a 14 year period.  Oversamples of military and poor white respondents were dropped after these stats were generated.
  
• For these statistics, only white non-Hispanic respondents were included.
  
• Only a few of the people in the sample took IQ tests in high school.  Of these, a median correlation of .81 was inferred between IQ tests and Armed Services Vocational Aptitude test.  That is, in spite of your claim, that top number is not IQ but an estimated IQ inferred from the ASVA.  
  
• Your claim, " there is a researched table that shows the percentage of undesirable life outcomes occurring, relative to the IQ of the person" is false.
  
• This is a highly abridged table, showing selected results from a table intended to show that intelligence is a better predictor of individual outcomes over 14 years than socio-economic status.  Obviously, using the Armed Services test as means of quantifying intelligence introduces significant bias (nobody's looking at musical aptitude or sense of humor in an Armed Services aptitude test) and confining data to non-Hispanic whites obviously distorts the assessment of socio-economic status.
  
• The data is so old fashioned that to qualify as middle class you had to have a high school diploma, never been incarcerated, remain married to first wife and never give birth to a child out of wedlock.  Better than 40% of modern women have children out of wedlock but Herrnstein and Murray throw all of these women into the failed end of socio-economic outcomes.
  
• Super old fashioned, kind of racist, and not using IQ.
  

• You claim that  "Schmidt and Hunter (1990) examined the correlation between job performance and IQ" but that is false.
  
• But Hunter and Schmidt explain, "The hypothesis was tested that the standard deviation of employee output as a percentage of mean output  increases as a function of the complexity level of the job."  The paper broadly found that as a job grew more complex, a wider variety of output was delivered by employees.  Nothing about intelligence or IQ at all.  In fact, a quick search for terms "IQ" "Intel" and "Quotient" all came up zero.
  
• The table you snapshot and the data you describe are not in the paper you linked to.
  

• You claim that Huffcutt & Arthur found that IQ was the best predictor of job performance but that is false.
  
• Huffcutt & Arthur found that Hunter & Hunter (1984) had totally under-represented the value of the job interview and went on to conclude, "highly structured interviews provide essentially the same validity as ability tests"
  
• The paper concluded the opposite of what you claimed it concluded.
  
• Huffcutt & Arthur did not specify Intelligence as the ability tested, nor IQ as the test of Intelligence.
  

• You claim "with education, occupation and income levels, IQ consistently produced the best correlation." 
  
• We should note that more than a hundred different types of intelligence tests were used- which raises plenty of questions regarding the correlations between one kind of intelligence test and the next.
  
• This data claims that neither intelligence nor socio-economic index correlates very strongly to income.
  
•  I would think socio-economic status would correlate with income far more than claimed here.

• You claim "this heuristic guess correlates very well with IQ"
  
• but, in fact, Dennison only reported  "weakly correlated with objective intelligence"
  

• As you say, this study reduces the value of perceived intelligence because shy people's intelligence is consistently undervalued in the first few meeting and talkative people's intelligence is consistently overvalued in the first few meetings.  This weakens further Denisson's "weak correlation"

• Let's note that the Southern Poverty Law Center defines Gottfredson as a promoter of eugenicism, scientific racism, and white nationalism.
  
• Gottfredson has taken more than a quarter million dollars in grants from the White Supremacist Pioneer Fund to advance eugenics research.
  

• The table also appears in "The Bell Curve"
  

• Ultimately, this data comes from the  National Longitudinal Survey of Youth 1979 which began in 1979 with 12,686 men and women born in 1957-64 (ages 14–22).  Members were interviewed annually from '79 to '94 but this table was published in '94 so this data cover outcomes over a 14 year period.  Oversamples of military and poor white respondents were dropped after these stats were generated.
  
• For these statistics, only white non-Hispanic respondents were included.
  
• Only a few of the people in the sample took IQ tests in high school.  Of these, a median correlation of .81 was inferred between IQ tests and Armed Services Vocational Aptitude test.  That is, in spite of your claim, that top number is not IQ but an estimated IQ inferred from the ASVA.  
  

• Your claim, " there is a researched table that shows the percentage of undesirable life outcomes occurring, relative to the IQ of the person" is false.
  
• This is a highly abridged table, showing selected results from a table intended to show that intelligence is a better predictor of individual outcomes over 14 years than socio-economic status.  Obviously, using the Armed Services test as means of quantifying intelligence introduces significant bias (nobody's looking at musical aptitude or sense of humor in an Armed Services aptitude test) and confining data to non-Hispanic whites obviously distorts the assessment of socio-economic status.
  
• The data is so old fashioned that to qualify as middle class you had to have a high school diploma, never been incarcerated, remain married to first wife and never give birth to a child out of wedlock.  Better than 40% of modern women have children out of wedlock but Herrnstein and Murray throw all of these women into the failed end of socio-economic outcomes.
  
• Super old fashioned, kind of racist, and not using IQ.
  

• You claim that  "Schmidt and Hunter (1990) examined the correlation between job performance and IQ" but that is false.
  
• But Hunter and Schmidt explain, "The hypothesis was tested that the standard deviation of employee output as a percentage of mean output  increases as a function of the complexity level of the job."  The paper broadly found that as a job grew more complex, a wider variety of output was delivered by employees.  Nothing about intelligence or IQ at all.  In fact, a quick search for terms "IQ" "Intel" and "Quotient" all came up zero.
  
• The table you snapshot and the data you describe are not in the paper you linked to.
  

 Huffcutt & Arthur (1994) Interview.pdf (radford.edu) 

• You claim that Huffcutt & Arthur found that IQ was the best predictor of job performance but that is false.
  
• Huffcutt & Arthur found that Hunter & Hunter (1984) had totally under-represented the value of the job interview and went on to conclude, "highly structured interviews provide essentially the same validity as ability tests"
  
• The paper concluded the opposite of what you claimed it concluded.
  
• Huffcutt & Arthur did not specify Intelligence as the ability tested, nor IQ as the test of Intelligence.
  

doi:10.1016/j.intell.2006.09.004 (emilkirkegaard.dk) 

• You claim "with education, occupation and income levels, IQ consistently produced the best correlation." 
  
• We should note that more than a hundred different types of intelligence tests were used- which raises plenty of questions regarding the correlations between one kind of intelligence test and the next.
  
• This data claims that neither intelligence nor socio-economic index correlates very strongly to income.
  
•  I would think socio-economic status would correlate with income far more than claimed here.
  

Antecedents and consequences of peer‐rated intelligence - Denissen - 2011 - European Journal of Personality - Wiley Online Library

• You claim "this heuristic guess correlates very well with IQ"
  
• but, in fact, Dennison only reported  "weakly correlated with objective intelligence"
  

(PDF) Perceptions of Intelligence in Leaderless Groups: The Dynamic Effects of Shyness and Acquaintance (researchgate.net) .

• As you say, this study reduces the value of perceived intelligence because shy people's intelligence is consistently undervalued in the first few meeting and talkative people's intelligence is consistently overvalued in the first few meetings.  This weakens further Denisson's "weak correlation"
  

I think the results are repeatable enough to conclude that IQ measures some aspects of the human intellect but hardly encompasses what humans mean when we say intelligence. 

According to the data I provided, that isn't true. That's why a heuristic evaluation of people's intelligence correlated really well with a formal I.Q. test.

We were just recently discussing Bobby Fisher's high IQ although that sad bastard was an unemployable schizophrenic who pissed himself to death because he wasn't smart enough to go to a doctor.  Was Bobby Fisher therefore intelligent?  I know a fair number of brilliant manic depressives but the people I know with a real genius for living an admirable life would probably score about 100 on that test.

You love anecdotes lol.

I think IQ is over-rated and fairly slanted towards a 19th century European male model of success- lawyers and doctors and engineers.  What of empathy?  What of self-awareness? What of leadership and problem-solving?  What of wit and humor- clearly signs of intelligence  that no IQ test can capture.  What of instinct- which is often the smartest part of us though we don't understand how or why?   IQ might correlate with job performance because it favors those  intellects which dot their i's and cross their t's, but tidy thinkers are seldom inspired and the truly creative seldom score high in IQ.

General intelligence ('g', of which I.Q. is a proxy for) has built-in a calculation of these facets.   In other words, general intelligence does not reject the notion that things like "empathy" and "humor" (to some degree) are part of general intelligence.   So, in essence, I agree with you that I.Q. should factor these attributes into it, and it already does to some degree.  The reason why it only factors these attributes "to some degree" is because a lot of those traits you listed can be learned. Clearly, "leadership" can be taught to people to some degree, and whilst you could argue that some people have a natural affinity for leadership (I agree), the environment will play a large part in a person's ability to acquire leadership skills (and thus make them better at leadership).

It makes no sense when assessing I.Q., which attempts to assess genetic general intelligence, to have factors which are purely environment factors.

In other words, testing "leadership skills" isn't going to provide us a strong correlate for raw general intelligence because it is muddied too much by environmental factors, defeating the purpose of measuring general intelligence.

Instead, it is better to test for general intelligence and let the genetic components of "humor", "leadership", "empathy" etc. be naturally integrated into the general intelligence assessment.

We can recognize that dogs are smarter  than sheep but an IQ test can't tell us why.  Or why some dogs are smarter than others.  Clearly adults are smarter than children in most ways so intelligence changes and adapts as we grow and age.   Obviously, there's more to intelligence than can be expressed in letters and numbers and in fact those abstractions came along only recently, long after humans had distinguished themselves as more intelligent than any other species.

You're moving the goalposts when you demand of I.Q. to explain "why". I.Q. is designed to be a proxy for general intelligence. That's it. Is it a valid metric in that regard? I've argued yes.

We see general intelligence proxy tests (I.Q. tests) and see that some people perform better than others. Does that mean we shouldn't use I.Q. tests?

Overall, I'm IQ skeptical and my experience tells me (hey, what about experience?)  that applying numbers to people's capacity is often self-defeating or at least self-limiting.  Tell people that a 4 minute mile is the limit of human capability and for some reason everybody believes it until one day that record is broken and suddenly thousands more find that ability within themselves.

Applying numbers (based on I.Q. results) to people's capacity produces moderate correlations to desirable life outcomes.

It's also often the best metric for predicting performance.  Clearly, I.Q. is having some impact on performance.

I agree that if you inspire people in various ways, you'll have them performing better. But you'll never have 55 I.Q. Pygmies becoming competent neuro-scientists at Yale, and you'll frustrate everyone involved if you force it.

But then when I ask for a definition of INTELLIGENCE you give me the definition for GENERAL INTELLIGENCE.  I say that IQ doesn't cover all of what we recognize as  intelligence.

• You disagree with me by saying that perception correlates well  with test outcomes but in a later post agree that Dennison found that the correlation was weak for starters and only improved over time.  Dennison only claims correlation to objective intelligence, which more or less agrees with my contention.
  
• You agree with me by limiting your definition of INTELLIGENCE to the smaller subset, GENERAL INTELLIGENCE.  
  

My contention was only that GENERAL INTELLIGENCE or objective intelligence is not the same thing as intelligence generally.  Huffcutt & Arthur would seem to agree with my point as does Dennison, as do you at some points in spite of your objection here.

Integrating multiple cognitive abilities in the service of solving a novel problem is also learned. Understimulated children often demonstrate cognitive deficits.

I don't think most would agree with your assertion that IQ tests are only or even mostly interested in the heritability of general intelligence.  Yes, that is the primary interest of the eugenicists you cite, which makes sense because their mission is to justify their assertions of political supremacy with some kind of objective measure of genetic superiority.  However, the general figure for the heritability of IQ, according to the APA is 0.45 for children, and rises to around 0.75 for late adolescents and adults.  Heritability measures in infancy are as low as 0.2, around 0.4 in middle childhood, and as high as 0.9 in adulthood.   That is, stripped of influences of environment, genetics only show a weak correlation to IQ.  I think its fair to say that most psychometrists, absent those tainted by political agenda,  are trying to assess the whole picture, not genetic general intelligence alone but also the influence of environment and the relationship between the two.

I think you have circled around to agreeing with me.  IQ can't capture all of what we call INTELLIGENCE including things like leaderships skills.

I can see that some important qualities might be reflected by IQ.  I have never seen an IQ test variant that quantified humor or empathy.  I don't agree that such qualities must necessarily or naturally flow from general intelligence.

I'm not demanding an explanation, I'm illustrating the limitations of objective intelligence assessment. We are able to perceive differences in intelligence that we have no hope of quantifying within the limited scope of verbal measurement.

You'd have to define performance more specifically.  The bell curvers were defining performance as holding a job for men and bearing children within wedlock for the ladies.  Kirkegaard's survey was looking at education level, occupation level (*), and income level as measures of performance.  I think IQ probably is a better or best predictor for educational, occupational, and financial attainment but again, all those measures fall within a certain Western paradigm.  We want to be careful not to mistake such metrics as a measure of human worth or an indicator of superior ability.

*I do wonder how "occupational levels" were measured and what biases revealed therein.

So, there's is a racial component to your assertions.  I don't know how you could possibly make such an assertion while failing to control for Western values.  Your definition of general intelligence may correlate with educational, occupational, and financial attainment but I assume that Bushmen are just as able to demonstrate their fluid ability to integrate multiple cognitive abilities in the service of solving problems within the context of the Kalahari as we are in the context of some Western metropolis.  I likewise assume that if the Mbuti were running the tests for cognitive abilities, big city white boys might take a significant hit in the scoring.  

 I may not be able to find a specific example of a Mbuti  neuroscientist, but the stated avg IQ score across all of sub-Saharan Africa is only 71 and nevertheless we see some prestigious neuroscientists excelling in Western cultures.  I do assume that some excellent individual Mbuti are entirely capable of such achievement given the right environment but if I'm reading you right, you don't. 

• Let's note that the Southern Poverty Law Center defines Gottfredson as a promoter of eugenicism, scientific racism, and white nationalism.
  
• Gottfredson has taken more than a quarter million dollars in grants from the White Supremacist Pioneer Fund to advance eugenics research.
  

These notes are both Ad Hominem attacks and are thus logically invalid.

• The table also appears in "The Bell Curve"
  

Your point?

For your second dot-point, we should note that I.Q. only correlates with itself at 0.87. Still, the researcher's  estimates have this 0.06 gap because ASVA is 0.06 less correlate than I.Q. is for I.Q, so you're right in arguing that this is a limitation of the study. However, an unaccounted 0.06 correlation effect is quite small and we should regard the impact it could have as such.

The reason why it only factors these attributes "to some degree" is because a lot of those traits you listed can be learned. ..... let the genetic components of "humor", "leadership", "empathy" etc. be naturally integrated into the general intelligence assessment.

For you second dot point, the data remains valid because the cultural snapshot applied to *all* participants at the time. Effectively, controls for environment. I'd also argue that, "high school diploma, never been incarcerated, remain married to first wife and never give birth to a child out of wedlock" are all positive life outcomes, hence the data remains valid. 40% of modern women, I would argue, are likely to have lower I.Q. for making a bad decision (child out of wedlock). Albeit, this point probably requires its own thread (I might make one). Hence, this data remains valid, in regards to I.Q.

Your final dot point is mostly Ad Hominem against the data. "Not using IQ" is addressed above.

• You claim that  "Schmidt and Hunter (1990) examined the correlation between job performance and IQ" but that is false.
  
• But Hunter and Schmidt explain, "The hypothesis was tested that the standard deviation of employee output as a percentage of mean output  increases as a function of the complexity level of the job."  The paper broadly found that as a job grew more complex, a wider variety of output was delivered by employees.  Nothing about intelligence or IQ at all.  In fact, a quick search for terms "IQ" "Intel" and "Quotient" all came up zero.
  
• The table you snapshot and the data you describe are not in the paper you linked to.
  

So Hunter and Schmidt use the term "mental ability" and "cognitive ability" instead of "IQ". It's fair enough to say that I should have noted this in the OP.

The table I snapshotted was not directly from the study, but rather a compilation of data from the study. For example, Hunter and Schmidt refer to mental ability specifically here:

"On the basis of Hunter's (1980) findings, the true-score correlations for general mental ability and supervisory ratings are .65 for high-complexity jobs, .57 for medium-complexity jobs, and .44 for low complexity jobs (when complexity is as defined in the text and in Footnote I)." 

The average of: 0.65+0.57+0.44 = 0.5533 (to 4 d.p.), but the data they used had a higher percentage of medium complexity and low complexity jobs, and so the average of the data they used makes "mental ability" correlate at roughly 0.51 in their study (as depicted in the table).

 Huffcutt & Arthur (1994) Interview.pdf (radford.edu) 

• You claim that Huffcutt & Arthur found that IQ was the best predictor of job performance but that is false.
  
• Huffcutt & Arthur found that Hunter & Hunter (1984) had totally under-represented the value of the job interview and went on to conclude, "highly structured interviews provide essentially the same validity as ability tests"
  
• The paper concluded the opposite of what you claimed it concluded.
  
• Huffcutt & Arthur did not specify Intelligence as the ability tested, nor IQ as the test of Intelligence.
  

Claiming that "highly structured interviews provide essentially the same validity as ability tests" doesn't detract from the validity of I.Q as a metric.

It's worth noting that intelligence/I.Q. is denoted as "Ability Composite" in this paper (which is admittedly a strange way to say intelligence).

doi:10.1016/j.intell.2006.09.004 (emilkirkegaard.dk) 

• You claim "with education, occupation and income levels, IQ consistently produced the best correlation." 
  
• We should note that more than a hundred different types of intelligence tests were used- which raises plenty of questions regarding the correlations between one kind of intelligence test and the next.
  
• This data claims that neither intelligence nor socio-economic index correlates very strongly to income.
  
•  I would think socio-economic status would correlate with income far more than claimed here.
  

We don't need to know how well I.Q. tests correlate with each other. All we need to know is how well the I.Q. test correlates with general intelligence, and then we can take an average of all the I.Q. tests in question.

You can think your final dot-point if you like, but you don't have any data to back you up. I do.

Antecedents and consequences of peer‐rated intelligence - Denissen - 2011 - European Journal of Personality - Wiley Online Library

-->@oromagi

But then when I ask for a definition of INTELLIGENCE you give me the definition for GENERAL INTELLIGENCE.  I say that IQ doesn't cover all of what we recognize as  intelligence.

• You disagree with me by saying that perception correlates well  with test outcomes but in a later post agree that Dennison found that the correlation was weak for starters and only improved over time.  Dennison only claims correlation to objective intelligence, which more or less agrees with my contention.
  
• You agree with me by limiting your definition of INTELLIGENCE to the smaller subset, GENERAL INTELLIGENCE.  
  

My contention was only that GENERAL INTELLIGENCE or objective intelligence is not the same thing as intelligence generally.  Huffcutt & Arthur would seem to agree with my point as does Dennison, as do you at some points in spite of your objection here.

I generally agree with what you're saying here.

As for "IQ doesn't cover all of what we recognize as  intelligence", "I.Q." not covering what is heuristically understood as intelligence does not mean we should make it match the heuristic understanding.

It would be like giving an English test to an African tribe, one of which has never seen English before, and concluding that they have zero I.Q. because they couldn't answer anything. Clearly, you need to control for learned knowledge. Likewise, "leadership skills", "humor" etc. suffer, to some degree, from this problem, too.

I don't think most would agree with your assertion that IQ tests are only or even mostly interested in the heritability of general intelligence.  Yes, that is the primary interest of the eugenicists you cite, which makes sense because their mission is to justify their assertions of political supremacy with some kind of objective measure of genetic superiority.  However, the general figure for the heritability of IQ, according to the APA is 0.45 for children, and rises to around 0.75 for late adolescents and adults.  Heritability measures in infancy are as low as 0.2, around 0.4 in middle childhood, and as high as 0.9 in adulthood.   That is, stripped of influences of environment, genetics only show a weak correlation to IQ.  I think its fair to say that most psychometrists, absent those tainted by political agenda,  are trying to assess the whole picture, not genetic general intelligence alone but also the influence of environment and the relationship between the two.

Firstly, stop Ad homming people.  

Secondly, I.Q. tests by function do assess the genetic component of I.Q. That might not be the primary interest or even an interest of administering the test,

Thirdly, I actually pretty much agree with your quoted estimations of I.Q. heritability. 

I think you have circled around to agreeing with me.  IQ can't capture all of what we call INTELLIGENCE including things like leaderships skills.

 I agree that I.Q. doesn't specifically measure for "leadership skills", measurement of I.Q. would be *worse* if it did.  

I can see that some important qualities might be reflected by IQ.  I have never seen an IQ test variant that quantified humor or empathy.  I don't agree that such qualities must necessarily or naturally flow from general intelligence.

I don't know how many ways I can make this point, but we'll see:

If I attended multiple seminars for humor analysis, if I watched sitcoms daily, if I spent time in a bar testing my stand-up routine on people, would not those experiences influence my ability to produce humor? Would I not have an advantage in learned knowledge if we tested for humor, especially against someone who didn't do these things? 

Do you see the difference between "learned knowledge" and general intelligence?

I'm not demanding an explanation, I'm illustrating the limitations of objective intelligence assessment. We are able to perceive differences in intelligence that we have no hope of quantifying within the limited scope of verbal measurement.

We don't need to answer this "why" in order for I.Q. to be a valid metric. The limitation you illustrated isn't relevant to the thread.

You'd have to define performance more specifically.  The bell curvers were defining performance as holding a job for men and bearing children within wedlock for the ladies.  Kirkegaard's survey was looking at education level, occupation level (*), and income level as measures of performance.  I think IQ probably is a better or best predictor for educational, occupational, and financial attainment but again, all those measures fall within a certain Western paradigm.  We want to be careful not to mistake such metrics as a measure of human worth or an indicator of superior ability.

*I do wonder how "occupational levels" were measured and what biases revealed therein.

By performance, I mean ability to perform at work, ability to make income and ability to become educated.

I also think that bearing children within wedlock is a desirable life outcome for ladies (and society). However, this is absolutely debatable (and probably requires its own thread).

So, one of the in-built premises of my argument is that civilizations are desirable.

I have used a racial example but that is not part of my "assertions". You can agree that I.Q. is a valid metric without believing that human races exist.

I don't particularly know about Pygmies, but I have done a small amount of research on Aboriginal Australians. They actually have a superior level of spatial awareness. If we were to use the same distribution as I.Q. test data presents, Aboriginal Australians have 119 points, which puts them above an entire standard deviation of the global average (can't find my source on this though). If you were to set-up a test that measured spatial awareness, Aboriginal Australians would almost certainly do the best. However, that doesn't mean they are intelligent or that it makes sense to measure intelligence through spatial awareness levels.

Clearly, if you find the Mbuti people who are five standard deviations (1 / 1744278) above their race's I.Q (you couldn't even find them lol), they are intelligent enough to become competent neuroscientists. This is not surprising. This doesn't meant that all Mbuti people are capable of becoming competent neuroscientists.

don't know why you've decided to argue now that the environment could make for this gigantic gap in I.Q,

I saw Joe Rogan’s test score on his Instagram and decided to see how I stack up

So I took the online IQ test at test-iq.org. 

It said:

Impressive! You completed the test in: 16 m 36 sThat's faster than 97% of people tested.Your strongest category is Pattern Recognition where you scored higher than 99% of people tested.

I didn't want to pay the $19.99, so I didn't get the actual score.

Added07.31.21 12:05PM

-->@oromagi

After they asked the 20 questions.

-->@oromagi

I don't know. Joe Rogan has an IQ of 127 based off his test results from the BMI Certified IQ Test he took. Maybe they are just capitalists?

Added07.31.21 01:10PM

-->@oromagi

The test claims to be the most accurate online IQ test in the United States. Although there is a free Demo IQ test on the site, there is a more in-depth Certified IQ Test that comes with a $20 fee.

WIKI: Valid ad hominem arguments occur in informal logic, where the person making the argument relies on arguments from authority such as testimony, expertise, or on a selective presentation of information supporting the position they are advocating. In this case, counter-arguments may be made that the target is dishonest, lacks the claimed expertise, or has a conflict of interest.

...

What people?  Eugenicists?  Is it an ad hom to call a Eugenicist a Eugenicist?   Are you denying that taking money from parties with a political interest in a particular outcome exposes the data to skepticism when that particular outcome is published?  If a study stating that smoking is good for you comes out and I complain that the study was sponsored by RJ Reynolds, is that mere insult or is that a legitimate cause for skepticism? 

Out of labor for 1 month is one bad life out indicator.  The original data is so old-fashioned that they only counted men, expecting women to drop out of the labor force regular for birthing and raising children.  I'd argue that if our priorities were straight, all parents would drop of the labor force for at least month as part of bonding with infants and other childcare responsibilities.

Had an illegitimate child: is another bad life indicator.  I understand why white supremacists are very interested in controlling breeding for the purpose of enforcing racial purity but 40% of all US mothers now have children deemed "illegitimate" by the old-fashioned standards of the early '80's.  Are we really still calling this a bad outcome.  Naturally, this bad outcome is only counting women as responsible for children born out wedlock.  Fathering illegitimate children is not considered a  similarly negative outcome, apparently.  [...] Again, better than 40% of US  women are having babies without a marriage license these days.  Are you really lumping all those women into a bad life outcome.

Lives in poverty: There's the classic indicator of classist bias, measuring positive life outcomes according to money made. Many people choose low-paying jobs out of a sense of civic duty- teachers, nurses, military, clergy, social workers, etc or because their priorities supersede financial considerations- artists, athletes, farmers, etc.  Many highly intelligent people choose poverty, experience contentment, yet end up the bad outcomes list.

[Schmidt and Hunter point] Yes, "mental ability" seems to claim a much wider range than "general intelligence"

Valid but not "best."  You claimed Huffcutt & Arthur supported "best predictor" but Huffcutt & Arthur said that Hunter (1980) badly underestimated the value of the job interview as the predictor of job performance and that when the job interview was highly structured surpassed IQ as a predictor.

You seem to be assuming that Huffcutt's "ability composite "represents IQ but I don't see where Huffcutt made that connection.