I.Q. Validity

Author: Analgesic.Spectre ,

Topic's posts

Posts in total: 44
  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    A plethora of times I've seen words to the effect of, "IQ isn't necessarily an accurate indicator of intelligence" (although that particular quote is far less egregious than I've seen), implying that I.Q. is often inaccurate in measuring intelligence. Rather than individually addressing instances of said claims, I'll expound upon the topic here in a holistic way.

    Generally, I.Q. tests are a combination of measuring: pattern recognition, verbal comprehension, mathematical ability, vocabulary and short-term memory. Whilst these are not perfect measurements of intelligence, they are heavily correlated with predictions of how well peers will rate a person's intelligence, school and workplace abilities:  (https://onlinelibrary.wiley.com/doi/abs/10.1002/per.799 ; https://www.tandfonline.com/doi/abs/10.1080/00221325.1979.10533422 ;  http://www.ingentaconnect.com/content/sbp/sbp/1977/00000005/00000001/art00016). Unless people's perception of intelligence is wild guessing, it matches measurements of I.Q. exceedingly well.

    In fact, I.Q. predicts income and educational achievement (things which, I hope we can agree, are indicators of intelligence) better than parental socioeconomic status (http://www.emilkirkegaard.dk/en/wp-content/uploads/Intelligence-and-socioeconomic-success-A-meta-analytic-review-of-longitudinal-research.pdf). Not only that, but I.Q. is the best predictor of educational level, occupational level and income level (again, more indications of intelligence). Surprisingly, I.Q. even beats 'grades' as a predictor of educational level. The average sample size for the groups is approximately 97,000: (http://thealternativehypothesis.org/index.php/2016/04/15/the-validity-of-iq/). 

    Whilst there isn't a panacea to alleviate the concern of I.Q. being intelligence, there is an abundance of research to suggest that I.Q. probably measures intelligence.
  • keithprosser
    keithprosser avatar
    Debates: 0
    Forum posts: 3,289
    2
    3
    3
    keithprosser avatar
    keithprosser
    I suppose we can accept that all other things being equal a 'more intelligent' person will score higher on an IQ test than someone 'less intelligent'.  

    The problem comes when all other things are not equal!

    Ideally, a IQ test would measure something innate about an individual, ie something that is indepedent of culture and education, but current test fail to to that as is demonstrated by the 'Flynn Effect', ie the observation that mean IQ scores have risen over time.   Unless people now are inherently smarter than their parents who were smarter than their parents (unlikely!) then IQ scores is not measuring an innate, culturally and educationally independent quality.   What eactly IQ measures is a matter of hot debate!

    This often comes up in the context of race.  The sad fact is that there are inequalities in human societies(such as the quality of schooling) that impact on IQ scores.  It's only when those social inequalities are erased that IQ scores might mean anything.
  • Swagnarok
    Swagnarok avatar
    Debates: 3
    Forum posts: 687
    2
    2
    5
    Swagnarok avatar
    Swagnarok
    The brain is a muscle. Like with actual muscles, some people got the genes for a more "athletic build" than others, while some people are predisposed to be less fit.
    However, it is my belief that hard work can overcome any such barriers, within reasonable limits. I would go so far as to say that even a guy with Down Syndrome could eventually build himself up to the point of quickly performing advanced mathematical calculations or computer programming, should he/she be willing to put himself through the sheer hell needed to develop his or her mind to that fine point. That's to say, it's all about using what you've got as efficiently as possible.
    That is to say, it's harder but not impossible. Of course, people don't really like to work extremely hard, so most, perhaps the overwhelming majority of people with lower IQs end up living like one would normally expect from dumb people. Oftentimes the only hope for these people (perhaps including myself) is that somebody comes along and makes them put in the work needed to perform highly.
  • drafterman
    drafterman avatar
    Debates: 6
    Forum posts: 4,706
    3
    6
    9
    drafterman avatar
    drafterman
    --> @Analgesic.Spectre
    In fact, I.Q. predicts income and educational achievement (things which, I hope we can agree, are indicators of intelligence)

    Not only that, but I.Q. is the best predictor of educational level, occupational level and income level (again, more indications of intelligence).
    And if we don't agree that those are indicators of intelligence?

  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @keithprosser
    I suppose we can accept that all other things being equal a 'more intelligent' person will score higher on an IQ test than someone 'less intelligent'. 
    Of course.

    Ideally, a IQ test would measure something innate about an individual, ie something that is indepedent of culture and education, but current test fail to to that as is demonstrated by the 'Flynn Effect', ie the observation that mean IQ scores have risen over time.
    Even if I.Q. was 100% determined by the environment (lol), I.Q. could be a valid measurement of intelligence. In actuality, your talking point referencing the Flynn Effect doesn't belong here, but I'll address it anyway because the point you make is still wrong.

    The Flynn Effect and differences between racial I.Qs have nothing to do with each other. James Flynn, the man who discovered the Flynn Effect, said this:

    "The magnitude of white/black IQ differences on Wechsler subtests at any given time iscorrelated with the g loadings of the subtests; the magnitudeof IQ gains over time on subtests is not usually so correlated;the causes of the two phenomena are not the same." (https://lesacreduprintemps19.files.wordpress.com/2013/08/the-e2809cflynn-effecte2809d-and-flynns-paradox.pdf).

    To put it very basically, whilst both White and Black I.Qs are increasing, the White-Black I.Q gap is not affected by this.

    IQ scores is not measuring an innate, culturally and educationally independent quality. 
    There is certainly an environmental aspect to I.Q. However, I.Q. is still a valid term, regardless of why there are I.Q. differences.

    The sad fact is that there are inequalities in human societies(such as the quality of schooling) that impact on IQ scores.
    This is like saying that we can't approximately guess what time the sun will rise tomorrow, if we don't know the time of year. Sure, we can't be ultra-accurate with our prediction (due to changing orbital patterns -- analogous to the environment), but we can determine a range in which the sun will rise between (due to the inherent properties of sunrise -- analogous to genetics).

  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @Swagnarok
    The brain is a muscle. Like with actual muscles, some people got the genes for a more "athletic build" than others, while some people are predisposed to be less fit. However, it is my belief that hard work can overcome any such barriers, within reasonable limits.
    I think you're underestimating the genetic percentage of I.Q. Whilst the environmental facet would be more pronounced in areas with abject poverty (Indian slums, rural China etc.), and is more pronounced with young age, I.Q. is largely genetic.

    Without slugging through time-intensive heavy research, I can briefly state that the I.Q. heritability falls anywhere from 57-86%, based on Wikipedia (with my personal estimate being 80% -- I might make a thread about this) (https://en.wikipedia.org/wiki/Heritability_of_IQ). Taking even the least favourable estimate of 57%, more than half of your I.Q. is based on genetics, and this is a barrier that certainly can not be overcome with "hard work".
  • keithprosser
    keithprosser avatar
    Debates: 0
    Forum posts: 3,289
    2
    3
    3
    keithprosser avatar
    keithprosser
    --> @Analgesic.Spectre
    My dear chap, I don't know if there are innate genetic differences between the races and neither do you.  But we do know that there are social factors that come into play.  Obviously if you want to investigate innate racial differences you have to eliminate all the non-innate, social and environmental factors first.or you won't know if you are measuring cause or effect.



    l

  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @keithprosser
    So, you've dropped most of my counter-points. That's a great start.

    My dear chap
    I'm not a chap, but whatever.

    I don't know if there are innate genetic differences between the races and neither do you.
    How could you possibly conclude this?

    I mean, you're not only wrong (http://run-down.com/guests/je_black_athletes_p2.php -- one of a plethora of things I could cite), but the arrogance in which you assert this is profoundly appalling. Not to mention that racial I.Q. differences isn't relevant to I.Q. validity at all LOL. You made a whole post making an enormous digression about racial I.Q, and you couldn't even use a single bit of research.

    Why is this site plagued with idiots masquerading as pseudo-intellectuals? Seriously, drafterman is repeatedly one (https://www.debateart.com/forum/topics/174?page=1&post_number=2 , and then consider his post here, just to name two), Smithereens is another (https://www.debateart.com/forum/topics/12?page=2&post_number=43 ; https://www.debateart.com/forum/topics/106?page=1&post_number=12 , Triangle.128k is another: https://www.debateart.com/forum/topics/187?page=1&post_number=21 ; https://www.debateart.com/forum/topics/187?page=1&post_number=22).

    What are these people doing here? This is a place to thoroughly examine claims, not to blurt out whatever pops into your head -- it's an incessant deluge of self-assured halfwits posting stream-of-thought.

    You are fantastically disappointing.

    Shoo.

  • keithprosser
    keithprosser avatar
    Debates: 0
    Forum posts: 3,289
    2
    3
    3
    keithprosser avatar
    keithprosser
    I did't make it clear that when I wrote "I don't know if there are innate genetic differences between the races and neither do you." I was referring only and specifically to genetic differences relating to IQ scores.

    if you and I both do an IQ test today it would reveal which of us has better solving power today.  I think it is obvious that how we do on such a test would depend on a whole raft of factors - it is not clear at all that the winner has better genes. 

    If we had taken the test as children it is quite likely the higher scorer would now be earning more in a better job than the loser - but it still wouldn't show that IQ score is genetically determined. 

    I do not dispute that IQ score measures something real about an individual - What is absurd is that racists use the 'average' scores to judge individuals.  Consider person X.  Whether he is 'ok' or not doesn't depend on his qualities but the average of which ever group he is classified as belonging to.

    if intelligence matters so much then logically we should discriminate between high-iq and low-iq individuals regardless of race.   But racists don't do that - they use the race average to excuse over-valuing low-iq members of their favoured race over high-iq members of non-favoured races. 

    It seems almost paradoxical that in this area it is the left that focuses on the individual and the right that treats people only as members of a collective.

    i m sorry to have drifted off iq tesing in the narrow.. in the narrow, i agree iq tests reveal real something about the individual beng tested.





     

     

  • Buddamoose
    Buddamoose avatar
    Debates: 4
    Forum posts: 3,177
    2
    3
    6
    Buddamoose avatar
    Buddamoose
    The issue most people have with IQ tests is that they misunderstand it as a measurement of innate intelligence independent of socio-economic factors. Its not, it just measure current intellitigence, and that IQ raises the more educated you become is thoroughly established. 

    Plus, i think this came up in another thread where I said that differences in average IQ are realatively meaningless because even though averages may differ, peak differentials are negligible. 

    Whereas say, peak strength differentials between women and men are hugely disproportionate to the tune of double to triple strength differentials if not more depending upon the type of excercise. 🤔

  • Smithereens
    Smithereens avatar
    Debates: 2
    Forum posts: 502
    2
    1
    4
    Smithereens avatar
    Smithereens
    IQ is a psychometric for g factor, which accounts for around 30-50% of variance between different cognitive skills. The other 50% variance is not accounted for by the g factor. If you claim g factor is a true measure of "intelligence" you're about 50% correct, which is exactly how much predictive validity you have to work with.

    Intelligence theories use factor models and IQ is one of them. the g factor is the most broad and is only apparent after dimentionality reduction. Each IQ test has subsets that all items load onto, and each loading itself loads onto the g factor with pretty high strength. As with all factor reductions, a lot of variance is sacrificed in the process. Anyone who claims the g factor is the only predictor of intelligence doesn't understand factor analysis. It's merely the most obvious predictor. A scree plot however would show you that the sum of the next dozen strongest factors summed together wouldn't match the eigenvalue of the g factor, so it's clearly the only factor worth using. 

    In short, IQ is a measure of g factor, g factor is the correlation of performance between unrelated cognitive tests, and the g factor accounts for up to 50% of the variance in performance. For an individual an IQ test result doesn't mean much, but in large populations we see trends and correlations that are very useful for scientific study. 
  • RationalMadman
    RationalMadman avatar
    Debates: 292
    Forum posts: 8,901
    10
    10
    11
    RationalMadman avatar
    RationalMadman
    IQ is how fast you learn, it doesn't measure retention, creativitiy or ability to coneg your thoughts.

    I am a genius of all four kinds.

  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    I did't make it clear that when I wrote "I don't know if there are innate genetic differences between the races and neither do you." I was referring only and specifically to genetic differences relating to IQ scores.
    My response to Swagnorok very briefly deals with the percentage of I.Q. that is genetic: https://www.debateart.com/forum/topics/194?page=1&post_number=6

    As for establishing the Black-White I.Q. gap, this meta-study should do the trick: http://humanvarieties.org/2013/01/15/100-years-of-testing-negro-intelligence/ 

    if you and I both do an IQ test today it would reveal which of us has better solving power today.  I think it is obvious that how we do on such a test would depend on a whole raft of factors - it is not clear at all that the winner has better genes.  
    I somewhat agree. If we live in different environments, the environmental factor will obfuscate genes to some degree. However, genetic I.Q. differences would be readily apparent, if we were to test Albert Einstein and a homie g from da hood.

    If we had taken the test as children it is quite likely the higher scorer would now be earning more in a better job than the loser - but it still wouldn't show that IQ score is genetically determined.  
    Actually, and this can be found on the Wikipedia page linked in the Swagnorok response, children's genetic component (heritability) of I.Q. is less detectable, when compared to late teens' and adults' (it's know as the Wilson Effect). For instance, if you're referring to 5 year olds, then the detectable heritability of I.Q. is roughly 22.25% (https://jaymans.files.wordpress.com/2014/04/iq-heritability-age.png).

    I do not dispute that IQ score measures something real about an individual - What is absurd is that racists use the 'average' scores to judge individuals.  Consider person X.  Whether he is 'ok' or not doesn't depend on his qualities but the average of which ever group he is classified as belonging to.
    Sure.

    if intelligence matters so much then logically we should discriminate between high-iq and low-iq individuals regardless of race.   But racists don't do that - they use the race average to excuse over-valuing low-iq members of their favoured race over high-iq members of non-favoured races.  
    Yes, there is a case for that.

    Also, please don't use the word racist. Reasons for my request can be found here: https://www.debateart.com/forum/topics/174

    It seems almost paradoxical that in this area it is the left that focuses on the individual and the right that treats people only as members of a collective.
    I don't care about identity politics.

    i m sorry to have drifted off iq tesing in the narrow.. in the narrow, i agree iq tests reveal real something about the individual beng tested.
    It's okay. If you want to talk about racial I.Q, make a thread on it.







  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @Buddamoose
    The issue most people have with IQ tests is that they misunderstand it as a measurement of innate intelligence independent of socio-economic factors. Its not, it just measure current intellitigence, and that IQ raises the more educated you become is thoroughly established. 
    I'd appreciate it if you could cite your research, so as to maintain the integrity of this thread.

    Plus, i think this came up in another thread where I said that differences in average IQ are realatively meaningless because even though averages may differ, peak differentials are negligible. 
    Are you suggesting that environmental factors are able to mostly compensate for heritability of intelligence?

    Whereas say, peak strength differentials between women and men are hugely disproportionate to the tune of double to triple strength differentials if not more depending upon the type of excercise.
    Women are a bit physically weaker (not that it matters too much in the modern world), but your proportions are off. Also, women can actually be stronger than men, in some instances.

    For example, in deadlifting at 105lbs, the 50-54 age range has the best man's lift at 220.4lbs (https://rawpowerlifting.com/wp-content/uploads/2018/08/World-Mens-Deadlift-Records.pdf), whilst the best woman's lift was 264.9lbs (https://rawpowerlifting.com/wp-content/uploads/2018/08/World-Womens-Deadlift-Records.pdf).

    Not bad for a weak and worthless woman, isn't it? ;)

  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @Smithereens
    IQ is a psychometric for g factor, which accounts for around 30-50% of variance between different cognitive skills. The other 50% variance is not accounted for by the g factor. If you claim g factor is a true measure of "intelligence" you're about 50% correct, which is exactly how much predictive validity you have to work with.

    Intelligence theories use factor models and IQ is one of them. the g factor is the most broad and is only apparent after dimentionality reduction. Each IQ test has subsets that all items load onto, and each loading itself loads onto the g factor with pretty high strength. As with all factor reductions, a lot of variance is sacrificed in the process. Anyone who claims the g factor is the only predictor of intelligence doesn't understand factor analysis. It's merely the most obvious predictor. A scree plot however would show you that the sum of the next dozen strongest factors summed together wouldn't match the eigenvalue of the g factor, so it's clearly the only factor worth using. 

    In short, IQ is a measure of g factor, g factor is the correlation of performance between unrelated cognitive tests, and the g factor accounts for up to 50% of the variance in performance. For an individual an IQ test result doesn't mean much, but in large populations we see trends and correlations that are very useful for scientific study. 

    > Makes many claims
    > Provides no sources

    The know-it-all kid is at it again.
  • RationalMadman
    RationalMadman avatar
    Debates: 292
    Forum posts: 8,901
    10
    10
    11
    RationalMadman avatar
    RationalMadman
    --> @Analgesic.Spectre
    I had a typo so I will repeat what I said in a clearer way:

    Intelligence has four components:

    1) How fast you spot connections and patterns. (highest form of this intelligence is your typical genius and usually war generals like Bonaparte and Khan)
    2) How much you retain from what you learned in one sitting from one, rapid input. ( highest form is the eidetic memory savant)
    3) Creativity in using what is known in a fast way to create extremely complex out of this world ideas at rates others can't (the 'mechanical' players in PvP games and people who prefer faster paced jobs and perform very well at them)
    4) Ability to convey complex ideas in ways non-geniuses can comprehend with high proficiency and ease (highest form is the typical politician etc)
  • Smithereens
    Smithereens avatar
    Debates: 2
    Forum posts: 502
    2
    1
    4
    Smithereens avatar
    Smithereens
    --> @Analgesic.Spectre
    "many claims" lol.

    With the exception of how much variance in performance g factor accounts for, the entirety of that post is just factor analysis explained. Asking for sources on how factor analysis works means nothing more than that you know absolutely nothing about how to perform factor analysis, and likely statistics in general. Go learn it before you start talking about real factor models, ignorant pup. 
  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @Smithereens
    "many claims" lol.

    With the exception of how much variance in performance g factor accounts for, the entirety of that post is just factor analysis explained. Asking for sources on how factor analysis works means nothing more than that you know absolutely nothing about how to perform factor analysis, and likely statistics in general. Go learn it before you start talking about real factor models, ignorant pup. 
    You have to be a special kind of know-it-all to admit that your post wasn't entirely an explanation of factor analysis, in that it has bare assertion(s), and then knowingly fail to provide evidence for your bare assertion(s).

    Your know-it-all-ism is amusing, as always.
  • Smithereens
    Smithereens avatar
    Debates: 2
    Forum posts: 502
    2
    1
    4
    Smithereens avatar
    Smithereens
    --> @Analgesic.Spectre
    it is all about factor analysis, with the exception that I pointed out. You're asking for sources on how to perform a mathematical operation, which is just displaying your illiteracy in the area you're trying to talk about. Obviously I'm not going to cite proof that maths works, you're simply going to have learn it yourself. The fact that you know so little that you don't even know what you don't know doesn't really surprise me.
  • Analgesic.Spectre
    Analgesic.Spectre avatar
    Debates: 1
    Forum posts: 474
    1
    1
    5
    Analgesic.Spectre avatar
    Analgesic.Spectre
    --> @Smithereens
    You're asking for sources on how to perform a mathematical operation
    Nope, I have no qualms with your factor analysis description. As you've already said, it would be silly to ask for sources on mere description. I have qualms with this:

    it is all about factor analysis, with the exception that I pointed out.

    You admitted that your whole post wasn't about factor analysis (I underlined the part for you), and then you knowingly failed to cite sources. I won't even get into the fact that I think you made several points which required sources, but you've already admitted that your whole post wasn't about factor analysis.

    Anyway, I'm done dealing with the blow-out of correcting your know-it-all ego. Either provide sources for the claims you admitted were unsourced, or be quiet.
  • Smithereens
    Smithereens avatar
    Debates: 2
    Forum posts: 502
    2
    1
    4
    Smithereens avatar
    Smithereens
    --> @Analgesic.Spectre
    With the exception of how much variance in performance g factor accounts for, the entirety of that post is just factor analysis explained.
    Go learn how to do a factor analysis and come back when you have an inkling of a clue about what the subject matter is lol. Your familiarity with this material is high school level. 

  • keithprosser
    keithprosser avatar
    Debates: 0
    Forum posts: 3,289
    2
    3
    3
    keithprosser avatar
    keithprosser
    Surely intelligence is all about the ability to solve problems and an IQ test is a set of problems to be solved.   Ergo, performance on a IQ test is going to correspond to problem-solving ability in a straight forward way.

    But obviously that isn't just the case with IQ tests - any set of problems could be used to test a person's ability to solve them.   The distinguishinig thing about IQ tests is that they are intended to be independent of acquired skills and examine a person's inherent, in-born 'intelligence'. 

    That is how IQ tests differ from regular 'exams', because exams usually focus on discovering how well a person has acqured some particiar kpwldge or skill.

    The issue, then, is how well (or badly) IQ tests work for revealing innate as opposed to acquired problem solving skills.   That is to say that IQ tests undoubtedly test and measure a persons problem solving intelligence.   What I dispute is that they measure innate ability reliably.





  • ethang5
    ethang5 avatar
    Debates: 1
    Forum posts: 4,457
    3
    3
    6
    ethang5 avatar
    ethang5
    --> @keithprosser

    What I dispute is that they measure innate ability reliably.
    I agree. But I'm interested in why you think so.
  • Smithereens
    Smithereens avatar
    Debates: 2
    Forum posts: 502
    2
    1
    4
    Smithereens avatar
    Smithereens
    --> @keithprosser
     What I dispute is that they measure innate ability reliably
    questioning validity is asking if they do measure innate cognitive ability. Questioning reliability is asking if they measure innate cognitive ability consistently. Despite the wording, a metric which is valid has to first be reliable. IQ tests are known to be valid, and they are stable across your entire life. So yes I would suggest they are reliable. 

  • keithprosser
    keithprosser avatar
    Debates: 0
    Forum posts: 3,289
    2
    3
    3
    keithprosser avatar
    keithprosser
    I think I'll stick with what I wrote, subject only to using 'innate' in the sense of 'genetically determined'.