Talk:Intelligence quotient/Archive 5

Archive 1

←

Archive 3

Archive 4

Archive 5

Archive 6

Archive 7

Archive 8

Table at the beginning of the article

There's a table at the beginning of the article where the IQ scores of a handful of pseudonymous people in different tests are reported. Firstly, the caption misrepresents the source by omitting the fact that these people are pre-adolescents among whom the reliability of IQ is lower. Secondly, I see no reason why such a table containing anecdotal data should be at the beginning of the article. In fact, I don't think the table should be in the article at all. If there's a need to discuss the reliability of IQ tests for children and their intercorrelations in the article, it should be done in a more scholarly and precise manner. Unless someone can justify the inclusion of this table, I will remove it.--Victor Chmara (talk) 19:18, 10 August 2010 (UTC)

To the contrary, the children come at just the age when child IQ tests are as reliable as they ever are. (In other words, you are mistaken when you say the caption misrepresents the source. You can check the source to be sure.) The author of the cited source is of course a world renowned expert on IQ testing, and he includes the data (in a different format) in his book to illustrate the point in the table caption. That author thinks that those data correct a substantial misconception about IQ testing that many people have, and I have every reason to believe that he is correct that that fact needs to be more widely known. I invite multiple Wikipedia editors to read the source and to discuss here how this well known fact (which could be referenced to other reliable secondary sources) fits into the article.

-- WeijiBaikeBianji (talk) 20:32, 10 August 2010 (UTC)

Only WISC is mentioned in the article body, so KABC-II and WJ-III need some words for the reader. PЄTЄRS J VЄСRUМВА ►TALK 21:13, 10 August 2010 (UTC)

I agree it would be helpful to identify the tests, and I'll look at the smoothest way to do that. (The caption and table column headings are things I experimented with in my sandbox, viewing sample page layouts with four different browsers, before I committed the edit to article space.) -- WeijiBaikeBianji (talk) 21:54, 10 August 2010 (UTC)

The table most certainly does not belong at the start of the article, it should go further down if anywhere. I don't feel the least bit of surprise at it in fact it seems to me to show that IQ scores are if anything a little more reproducible than I expected. I guess other people might find it interesting but it doesn't contribute very much to explaining the topic of the article or a summary of the article which is what the lead is about. Dmcq (talk) 22:44, 10 August 2010 (UTC)

WeijiBaikeBianji, the correlations between the columns of the table are 0.62, 0.68 and 0.81. The first two are, even if not implausible, clearly on the low side (when thinking of adult tests at least). The real problem, however, is that the table lacks context. When it is in the beginning of the article, it will be the first thing the reader will notice and perhaps the first thing he will learn about IQ tests in general, perhaps making him think from the outset that IQ tests are not reliable. Kaufman in his book certainly does not say that the table implies that IQ tests are not reliable. He discusses, for example, how different tests emphasize different abilities, how a good clinician must learn to reject scores that clearly do not reflect the testee's true abilities, and how small children in particular can be difficult to test reliably. (He does not seem to discuss, in this context, how the Flynn effect makes norms obsolete, or the fact that g factors derived from different test batteries are very highly correlated[1], both of which are important facts to know when interpreting the table.) The reader will not be able to interpret the table correctly if the issues it deals with are not explained in the article.

As an analogy to your table, what do you think would happen if someone added to the start of the much contested Race and intelligence article a chart reporting on racial differences in average brain size, justifying it by saying that the data correct a substantial misconception that many people have, and that the data need to be more widely known? Such an edit would be reverted in a microsecond, and rightly so, because, even if it is factually correct, it would be POV pushing to give it undue weight by placing it at the start of the article.

Your table should not be where it is now, and unfortunately I don't think there's a place for it anywhere in the article currently. Perhaps you could add a section discussing the differences between IQ tests, and the reliability of different tests. Your table might fit in such a context. However, even then the caption should mention that the people tested were small children, and that the table is about variability between different test batteries.

I will restore the bell curve pic[2] that used to be at the start of the article.--Victor Chmara (talk) 08:45, 11 August 2010 (UTC)

Agree with that. Another suitable picture I think might be one of the tests questions in commons:Category:intelligence which would illustrate the idea of intelligence testing. Dmcq (talk) 11:59, 11 August 2010 (UTC)

I'll look around at the images suitable for Wikimedia project licensing. I visited a foreign-language version of the Intelligence article (not this Intelligence quotient article) yesterday that had a beautiful image appropriate to that broader topic. As you can imagine, most foreign-language Wikipedias follow English Wikipedia in many matters of content, including images, but there is some variation here and there that may be food for thought for us. Thanks for the link to the category. -- WeijiBaikeBianji (talk) 13:30, 11 August 2010 (UTC)

Victor, you wrote, "perhaps making him think from the outset that IQ tests are not reliable." Which leads to the question, what does the article now say based on reliable, recent secondary sources about the reliability of IQ tests? After edit: I think Dmcq gave a rationale here (illustrations near the lede should illustrate the whole article) that similarly argues against having the bell curve image (which is a LOT less well sourced than the IQ score table) at the top of the article. The bell curve image is already well placed in the section about Wechsler and deviation scoring of IQ tests. -- WeijiBaikeBianji (talk) 13:23, 11 August 2010 (UTC)

I agree that there should be more discussion about reliability in the article. The bell curve image (which hardly needs more sources as it's completely uncontroversial) could as well be placed elsewhere in the article. Of the pics suggested by Dmcq, perhaps this one could be used at the start of the article:

--Victor Chmara (talk) 13:44, 11 August 2010 (UTC)

Great minds run in the same rut. I just put a Raven matrices-like image at the head of the article, and opened up a section on reliability, to which you and all editors are welcome to add sources. I think I said earlier on your user talk page that I have given very few of the longer intelligence-related articles thorough line-by-line readings, but I will now be reading this article with that degree of thoroughness and I'll be digging through three versions of a working paper I've presented in public seminars in the American Midwest and through new sources I have at hand for references and for interesting facts about IQ testing that are of interest to the general readership of Wikipedia. I next speak on the topic for a public audience in Minneapolis in early November. -- WeijiBaikeBianji (talk) 14:12, 11 August 2010 (UTC)

Just this one small change makes the article and topic much more approachable. :-) PЄTЄRS J VЄСRUМВА ►TALK 14:30, 11 August 2010 (UTC)

Since I was there, I was wondering about inserting the following (proposed bolded): "While the heritability of IQ has been investigated for nearly a century, controversy remains regarding the significance of individual and group heritability estimates,[4][5] and the mechanisms of inheritance are still a matter of some debate.[6]"

   Also on inheritance mechanisms being the subject of some debate, can that be stated in a way that indicates what the issue is? For example, not understand the genetic mechanism or some such? I'll defer to the more expert editors on wording, but even a summary should offer more in the way of information. PЄTЄRS J VЄСRUМВА ►TALK 14:41, 11 August 2010 (UTC)

Yes, the heritability section of this article needs to be extensively rewritten (I have some very good sources at hand, from the medical library and biology library of my state's flagship research university). Similarly, the article to which that section links also needs to be extensively rewritten, and perhaps merged with other articles and retitled as well. Right now the best and most current professional secondary sources on the subject are buried under older and less reliable primary or self-edited sources that really shouldn't be in the articles at all by WP:MEDRS. Now that I have waded into Wikipedia editing, I have a huge and ever growing do-list, but at least there are collaborators here who are curious about the literature and thus willing to read reliable sources on intelligence, genetics (a new source list I will be expanding considerably soon), and other related subjects. Thanks for raising issues that need to be reexamined by editors. -- WeijiBaikeBianji (talk) 16:13, 11 August 2010 (UTC)

Section edits coming up.

I saw an edit summary in a diff calling for more sources for the new reliability section of this article, and I expect to supply those soon (I hope as early as later today). On the basis of published sources all Wikipedians may consult I will also be restructuring this article, at first not adding or subtracting much content but simply moving sections together or reordering sections to match the usual treatment of the subject of IQ testing in published books and review articles. As that goes on, I will try also to update sources so that they rigorously meet Wikipedia source guidelines for medicine-related articles. If anything I'm doing looks weird or controversial, feel free to ask me about it here on the article talk page. And of course normal editing conditions still apply here. Other than that the article is on pending changes review, anyone can edit, and I encourage everyone else who watches the article or who surfs by to think about ways to improve and better source the article and to make it an even more valuable resource for readers than it already is. -- WeijiBaikeBianji (talk) 15:55, 15 August 2010 (UTC)

I will begin some of those section edits. At first I will just be moving text around without changing anything other than drop-dead-easy copyedits. -- WeijiBaikeBianji (talk) 20:59, 15 August 2010 (UTC)

I was distracted by the Arbitration Committee case on the Race and intelligence article recently, but I'll return now to doing section moves and eventually other edits on this article. As before, at first I'll just rearrange section order with matching the structure of this article to the structure of reliable secondary sources in mind. I won't be changing article text at first. Later on, I will be changing article text a lot—typically with notice to fellow editors here on the talk page—to add in sourced content from sources of the highest reliability, which is important for articles like this article that have medical and forensic implications that literally can be issues of life and death. -- WeijiBaikeBianji (talk) 22:06, 23 August 2010 (UTC)

I agree that the current structure of the article is a mess. Can you post a draft here of what you think the structure should be like? I also think we should create a separate article named History of intelligence testing or something like that, so that this article could concentrate on current results and controversies.--Victor Chmara (talk) 08:29, 24 August 2010 (UTC)

The history section doesn't seem large enough to warrant splitting off, so I can't see the point. I think invoking life and death over intelligence quotient is rather over the top. Dmcq (talk) 10:11, 24 August 2010 (UTC)

You haven't heard about the Atkins case? That's the court decision that literally makes IQ scores a matter of life or death, that is eligibility or not for capital punishment in the United States. It has been claimed in several articles on Wikipedia, including this one, that IQ has some implications for health, and that is why I agree with the suggestion of RexxS that it's best to apply the medical sourcing rules to articles on IQ. (Those rules strike the correct balance between primary and secondary sources, for one thing.)--?

At the very least the history section should have fewer than seven subsections.--Victor Chmara (talk) 10:26, 24 August 2010 (UTC)

Yes, that's an example of my not being bold so far. That's too fragmented a presentation of the history in its current form.--?

Off the top of my head, I think the structure should be something like this:

1. History

2. Test construction and reliability

3. General mental ability (g factor)

3. Validity and social significance

4. Environmental and genetic influences

5. Flynn effect

6. Group differences

7. Criticisms

8. In popular culture

--Victor Chmara (talk) 10:52, 24 August 2010 (UTC)

Thanks for the very concrete suggestions. I'll take your list of sections (by copy-and-paste) and annotate it here.

1. History - yes, in this article history should probably be first, and I agree with your separate suggestion that the history section shouldn't be so busted up into separate subsections

2. Test construction and reliability - I'm glad you added test construction here (I just found a book-length source on that over the weekend, not yet logged into my source list)

3. Validity and social significance - I think it's customary in most of the secondary literature to mention validity immediately after reliability (and "social significance" might be treated as one aspect of validity) You are, I suppose, thinking of some of the things Linda Gottfredson has written about IQ scores and life outcomes.

4. General mental ability (g factor) - noting that there is a separate article about General intelligence factor, and that most current tests are based on CHC theory, maybe this could be called Factor structure of tested abilities

5. Environmental and genetic influences - with of course "heritability" being the summary term that examines the balance of each, and with malleability/mutability/changeability definitely needing mention in that section

6. Flynn effect - I'll check secondary sources, as I think you are right that this usually follows the heritability section in a typical current source

7. Group differences - I think, by the way, that this would be a good neutral title for a rewritten version of the article that just went through the ArbCom case

8. Criticisms - this would draw in information from disciplines other than the psychometric subfield of psychology for NPOV

9. In popular culture - I think this kind of section is actually disfavored by veteran Wikipedia editors like John Broughton, although this is certainly a topic with more than a typical amount of popular culture references, and you certainly see sections like this all over Wikipedia

Thanks for the helpful suggestions. I've had to rearrange my whole office just to put the few dozen most useful sources nearest to my computer as I type. It will be good to dig into the sources together and rewrite the article. -- WeijiBaikeBianji (talk) 13:46, 24 August 2010 (UTC)

Good comments, thanks. "Factor structure of tested abilities" is a good heading. Flynn effect could also be rolled into the section on "Environmental and genetic influences". I agree that the Popular culture section is a bit superfluous, but I think high-IQ societies could perhaps be discussed there instead of in a section of their own (I think they're worth mentioning).

I don't think "Group differences in intelligence" is a good title for the R&I article for the reasons I discuss on the R&I talk page.--Victor Chmara (talk) 14:46, 24 August 2010 (UTC)

@ Victor and History of...: The current history section is very choppy and doesn't seem very cohesive in being a narrative on the evolution of testing. Perhaps simply rewriting that section would be a start, if it gets too big it can be summarized and spawn a daughter article. Also, unless I'm mistaken, shouldn't "1929" be "1939"? PЄTЄRS J VЄСRUМВА ►TALK 17:30, 26 August 2010 (UTC)

@Peters: I can probably redo the history section soon with material I have (footnoted and everything) from my working paper. @Dmcq: I'm glad to see you pondering the section levels, and the issues of what sections belong. I'll have to look at the included content of each section; I think validity would belong right after reliability (the typical order in secondary sources). -- WeijiBaikeBianji (talk) 22:16, 26 August 2010 (UTC)

To all: I've decided the most efficient way to do the edit of the history section, which is a good bit too long, is to copy the entire section off-line and edit it there. You'll see the results soon, with fewer sections and more up-to-date references. -- WeijiBaikeBianji (talk) 01:06, 28 August 2010 (UTC)

Group differences section now being actively edited.

I see from watching diffs that there is editor activity now in the group differences section of this article. As the section currently notes, the issue is very controversial. As many experienced Wikipedians know, the main article linked out to from that section was recently the subject of an Arbitration Committee case, which has just been decided. I have just reverted a graphic that was just kindly inserted into that section by an editor whom I have not had the pleasure of interacting with before. I am happy to discuss my rationale for reverting the graphic here on the article talk page. As a precaution and friendly reminder to new editors here, I will post at the top of this talk page a template that links to the ArbCom case decision. Let's discuss how we can improve that section and all sections of this important article. -- WeijiBaikeBianji (talk) 16:33, 26 August 2010 (UTC)

I see the image has been restored by the editor who submitted it at first. I think to avoid undue weight problems it would be necessary to mention a lot of criticism of the source from which the image comes, if the source is cited in such a brief paragraph that points to a longer article. It would be better editorial practice, based on my understanding of Wikipedia policy, to have no image at all there but perhaps a slightly expanded summary paragraph or two (cited to a balanced selection of current secondary literature) rather than relying on one source, and one source only, in that section. Thanks for any thoughts any of the rest of you have on this issue. -- WeijiBaikeBianji (talk) 19:53, 26 August 2010 (UTC)

I see it's (IQ by race, Lynn ete al. as I recall) gone again. It's definitely misplaced in this article, plain and simple. There's no value judgement associated with that. PЄTЄRS J VЄСRUМВА ►TALK 15:18, 27 August 2010 (UTC)

Group differences is a different topic from predictive validity. Now the article gives the impression that all research on the predictive validity of IQ was about group differences, whereas most research on validity is about individual differences within populations. We should have a small section on group differences, with links to Sex and intelligence and Race and intelligence, and a separate section called "Validity" or "Validity and social significance" with subsections on the associations between IQ and health, job performance, school performance, etc.--Victor Chmara (talk) 11:06, 31 August 2010 (UTC)

Other Intelligence Tests?

Besides IQ, what other forms of measuring cognitive abilities are there?

— Preceding unsigned comment added by Rhyming (talk • contribs) 09:36, 30 November 2008 (UTC)

Turkheimer et al.

A paper by Turkheimer et al. claiming that the heritability of IQ varies by SES in small children is discussed in the article. The paper itself is cited along with some non-scholarly website. This is problematic, as we should use reliable secondary sources. Citing this study by Turkheimer et al. is also problematic because it's just one study, and other studies[3][4] have failed to replicate its findings. I'm not sure if Turkheimer et al. should be discussed in this article at all, but if it is discussed, we should use secondary sources and point out that its results are contradicted by other studies and also perhaps make it clear in this context that the heritability of IQ rises with age.--Victor Chmara (talk) 10:56, 10 September 2010 (UTC)

Turkheimer has written some good recent review articles (thus, secondary sources) that well belong in this article as sources. He is, of course, one of the most experienced researchers in twin studies of human behavior. I agree with the general proposition that this article, and the several related articles, could be improved by deleting statements that can only be found in primary sources (especially unreplicated primary sources). In lieu of primary sources, we should all look for the latest reliable secondary sources for more statements that can be backed up by those (which may be the same statements already in the article, now cited to less reliable sources). A review of the best current secondary sources will also provide guidance for the overall structure of the article. The earlier discussion on this talk page had Victor, Dmcq, and I discussing a revision of the structure of this article that still seems sound to me. -- WeijiBaikeBianji (talk) 15:51, 10 September 2010 (UTC)

How many templates does one article need?

A new editor has kindly added a new template to the top of this article and to several other articles linked to from the template. So far he (and I, and the one other editor who has done any editing on that template) hasn't achieved a broad consensus about what the template is for, what should be included in it, and how it can help readers of Wikipedia. What do you think? -- WeijiBaikeBianji (talk) 00:24, 21 October 2010 (UTC)

What is wrong with having a template for the many intelligence articles? Obviously a template about intelligence articles will help reader interested in intelligence. That is why Wikipedia have templates for similar articles. There is no other intelligence template on this article and only one other template about "Human group differences" which is a much broader topic. A dispute regarding exactly what the template should contain is not a reason for removing it.Miradre (talk) 05:05, 21 October 2010 (UTC)

"How many roads must a template walk down, before you can call him a man...?"

The template is for human intelligence articles. It contains intelligence articles. Thanks WBB for adding an article to it, I hope others add stuff too. Woodsrock (talk) 12:19, 21 October 2010 (UTC)

Two paths forward

(1) A while ago, editors Victor Chmara and Dmcq and I discussed some restructuring of this article, with some rearrangements of sections. I still think that the outline largely proposed by Victor is sound and would be a good framework for improving this article. (2) Having been reminded about Wikipedia reliable source guidelines through editor discussion on other articles, I am beginning to flag article text statements cited only to primary research studies, to remind editors to collaborate in looking for reliable secondary sources for article text statements in this article and elsewhere. -- WeijiBaikeBianji (talk) 11:32, 26 October 2010 (UTC)

Again, primary sources are not disallowed. Explain concretely what is wrong with the specific sources you tag. Otherwise I will eventually remove the tags.Miradre (talk) 14:06, 26 October 2010 (UTC)

See the Wikipedia reliable sources guideline ("Articles should rely on secondary sources whenever possible. For example, a review article, monograph, or textbook is better than a primary research paper. When relying on primary sources, extreme caution is advised: Wikipedians should never interpret the content of primary sources for themselves. See Wikipedia:No original research.") and the Wikipedia guideline on reliable sources (medicine) ("Respect secondary sources Individual primary sources should not be cited or juxtaposed so as to "debunk" or contradict the conclusions of reliable secondary sources, unless the primary source itself directly makes such a claim (see Wikipedia:No original synthesis that advances a position). Controversies or areas of uncertainty in medicine should be illustrated with reliable secondary sources describing the varying viewpoints. The use and presentation of primary sources should also respect Wikipedia's policies on undue weight; that is, primary sources favoring a minority opinion should not be aggregated or presented devoid of context in such a way as to undermine proportionate representation of expert opinion in a field.") for guidance on what sources to prefer for Wikipedia article text, especially for statements about human intelligence or neuroscience or other medical claims. -- WeijiBaikeBianji (talk) 14:13, 26 October 2010 (UTC)

Again, primary sources are not disallowed generally. The science articles are full of peer-reviewed articles. If you have a specific problem with a specific source, then take it up here for discussion. But mass tagging peer-reviewed articles because they are peer-reviewed articles is not acceptable. Again, explain for each specific source what is the specific problem. Otherwise I will eventually remove the tags.Miradre (talk) 14:53, 26 October 2010 (UTC)

Miradre, you have referred on other article talk pages to what you observe in other articles. Wikipedia has 6,816,083 articles, and most of those have no article rating at all, or are rated as stubs or start class articles. The thing to do if you would rather use inductive learning (look at examples and then develop a practice for editing) rather than deductive learning (look at Wikipedia policies and guidelines and then develop a practice for editing) is to look at recently featured articles. Articles that have achieved featured article status and are about related topics may indeed be good examples of how sources are used. They have passed a review process that most Wikipedia articles have never experienced. One article I like, Confirmation bias, is an illustrative example. The article cites what are plainly primary research articles, supporting your statement that those can be allowable sources on Wikipedia. It also cites a rather larger number of standard textbooks and review articles from professional journals than most Wikipedia articles, which is a condition that I think more Wikipedia articles ought to achieve. Any statement in article text that is cited to a reliable secondary source doesn't have any sourcing issue, so a source tag is an opportunity for an editor to find a better source. Referring to secondary sources rather than primary sources, as the Wikipedia guidelines I quoted above note, helps ensure that primary sources used in articles are used with due weight, a separate issue from sourcing that relates to the core Wikipedia policy on neutral point of view. Again, the reason that Wikipedia editors have long had an editing template available to them to mark primary sources is that reliable sourcing is important for all of the 6,816,083 articles on Wikipedia, most of which do not have reliable sourcing yet. -- WeijiBaikeBianji (talk) 15:13, 26 October 2010 (UTC)

As a general point I will note that literature reviews in no way guarantees neutrality or accepted scholarly consensus. For example, regarding if genetics are a partial explanation for the racial gaps in the US regarding IQ and other achievement tests, both sides have produced literature reviews with totally opposing conclusions.

Looking at featured science articles, such as Big Bang, Coeliac disease, Ant, and the others at Wikipedia:Featured articles, there are numerous peer-reviewed articles as sources. Neither is there any general prohibition of such sources in any policy. Yes, there may be specific problems in specific situations. But you have not given any specific explanations for why each the sources you tagged have specific problems. Please explain the specific reasons for each source tagged or I will eventually remove the tags.Miradre (talk) 15:34, 26 October 2010 (UTC)

While secondary sources are of course preferable and articles should be mainly based on them, there is no reason to dogmatically oppose all use of primary sources. WP:PRIMARY lays out the rules for using primary sources:

Primary sources that have been reliably published may be used in Wikipedia, but only with care, because it is easy to misuse them. Any interpretation of primary source material requires a reliable secondary source for that interpretation. A primary source may only be used on Wikipedia to make straightforward, descriptive statements that any educated person, with access to the source but without specialist knowledge, will be able to verify are supported by the source. For example, an article about a novel may cite passages to describe the plot, but any interpretation needs a secondary source. Do not make analytic, synthetic, interpretive, explanatory, or evaluative claims about material found in a primary source. Do not base articles entirely on primary sources. Do not add unsourced material from your personal experience, because that would make Wikipedia a primary source of that material.

As to the structure of the article, this is what I propose, based on the earlier discussion:

1. History

2. Test construction and reliability

3. Factor structure of test batteries

4. Validity and social significance

5. Environmental and genetic influences

6. Flynn effect

7. Group differences

8. Criticisms

[9. In popular culture]

--Victor Chmara (talk) 16:23, 26 October 2010 (UTC)

Race and intelligence

An editor has been sticking a lrge amount about race and intelligence into this article. I've pointed out there's a separate article Race and intelligence but they are persisting.

The Race and intelligence article has been edited quite a lot recently. What I'm wondering is has it changed significantly so this article should be updated and how much about the subject should be here? I pointed out to the editor that the Sex section just before Race was quite small and the main stuff was also in another article. Dmcq (talk) 17:14, 5 September 2010 (UTC)

Good initial call. This article, which bears on its talk page a notice about the recent Arbitration Committee case, is surely of interest while editing several other articles that were enmeshed in that case. It's best to seek consensus on the talk page before making major changes of content in this article—as other editors and I have been careful to do. As we discuss, I encourage all editors to look up reliable sources on the subject and to suggest new sources that will be helpful to other editors. P.S. probably semiprotection of the main article (after a long period when it was full-protected) has driven some I.P. editors over here to do section edits. -- WeijiBaikeBianji (talk) 18:51, 5 September 2010 (UTC)

The pending changes protection has been removed from this page after the test period for it. Pity, I certainly thought it helped and I prefer it to the type of protection on Race and intelligence which stops all IP edits.

I'd have thought that the leader of the race and itelligence article would be the maximum one wold want here so anything not in that could usually be considered as being over he top here. Dmcq (talk) 19:49, 8 September 2010 (UTC)

Yes, in general if a set of Wikipedia articles have already established an article including subarticles structure, the whole point of the subarticle is to be the place to go on at somewhat greater length on a more specific topic. A good lede paragraph or two (which is something the editors over at the subarticle are still working on) ought to be about the right length for a similarly worded "For further information, see . . . " section over here. -- WeijiBaikeBianji (talk) 20:09, 8 September 2010 (UTC)

Surely in an article on I.Q in the section on I.Q. and race, one ought to mention Herrenstein and Murray's book "The Bell Curve," an exhaustive study on this very subject that show by examining scores of studies that race and I.Q are linked. These conclusions are attacked for being "controversial," (read "unpopular"), but the work is so thorough and so detailed it cannot simply be ignored because it offends some activists.Tholzel (talk) 01:10, 6 December 2010 (UTC)

There are many better books on the same subject. One of several I can readily suggest is Devlin, Bernie; Fienberg, Stephen E.; Resnick, Daniel P.; Roeder, Kathryn, eds. (1997). Intelligence, Genes, and Success: Scientists Respond to the Bell Curve. New York (NY): Springer. ISBN 978-0-387-94986-4. {{cite book}}: Invalid |ref=harv (help); Unknown parameter |coauthors= ignored (|author= suggested) (help); Unknown parameter |laydate= ignored (help); Unknown parameter |laysummary= ignored (help) There is a whole bibliography of books about IQ and related subjects to consider when deciding which sources to include in this article. -- WeijiBaikeBianji (talk, how I edit) 03:02, 8 December 2010 (UTC)

Correlation and causation

Correlation means a lot of things and all of these are described in Correlation and dependence. This includes that correlation does not imply causation. The source material says IQ correlates with something without making a statement about causation so why should "correlation" link to an article only about this principle? This leads the reader to the conclusion that this principle of correlation is uniquely relevant here, when the source doesn’t support that.Boothello (talk) 00:22, 11 November 2010 (UTC)

This question wouldn't mostly be directed to someone new here, as the article text has been as it is for a while, but what is the rationale for mentioning the IQ correlations at all? What does reliable secondary literature (rather than one or another researcher's unreplicated primary research finding) say about the correlations? Perhaps better sourcing is needed in the section of the article under discussion. -- WeijiBaikeBianji (talk, how I edit) 02:36, 11 November 2010 (UTC)

IQ tests were invented as performance predictors, so it makes perfect sense for the article to discuss whether they predict the things they were designed to predict. Maybe better sources can be found for this. In the meantime though, you aren't disputing the point that it makes no sense for the word "correlation" to link to an article about only one aspect of correlation, so I'm changing it back again.Boothello (talk) 03:53, 11 November 2010 (UTC)

IQ tests were invented to measure relative intelligence. With respect to predictability IQ predicts some things, and some things predict IQ. This general problem of direction of causality (which you might also be confusing) is exactly why some explanation of correlation vs. causation is relevant. But the extent to which it is discussed depends on reliable sources, and is probably more relevant to include in the body of the article rather than hidden under a link. aprock (talk) 04:05, 11 November 2010 (UTC)

According to the article about Alfred Binet, who invented intelligence tests, the tests were invented to predict scholastic performance and identify students who were in need of special education.Boothello (talk) 05:03, 11 November 2010 (UTC)

That's not what the article says at all: Binet made it his problem to establish the differences that separate the normal child from the abnormal, and to measure such differences. aprock (talk) 16:25, 11 November 2010 (UTC)

I will be adding a lot of see also references to Wikipedia articles not already linked from this article.

It happens that last year I gathered a bunch of Wikipedia article links for my own working paper project, and now I see that not all of those are yet linked from this article. So I will be updating the See Also section of this article to add those. WeijiBaikeBianji (talk) 15:37, 3 June 2010 (UTC)

Of course I ran into the issue that most of those articles need edits also. But as long as I have the Intelligence Citations list posted for all Wikipedians to refer to (and your suggestions of new sources for that are most welcome), it should be possible gradually to make progress in editing the various articles related to this topic. -- WeijiBaikeBianji (talk) 15:51, 28 July 2010 (UTC)

Great article. A nit on article links and "Psychometrics". As presented in the article, Psychometrics is (despite a link to a much better description in its own article), portrayed here as only associated with the sordid history of eugenics - an unintended smear on all those good folks in the field today, I'm sure, but an unjustified association nonetheless. Perhaps a following sentence to rectify that, or just linking to its better description in its own article, would be better. —Preceding unsigned comment added by 76.88.4.187 (talk) 18:23, 12 October 2010 (UTC)

I'm a little bit unclear about what part of the article text (as I surmise) you are referring to by your comment. (That's because I didn't write much of the article text here, and I am still becoming familiar with the article as a whole and with its details.) What sources would you suggest for a good overall view of psychometrics? -- WeijiBaikeBianji (talk) 23:05, 12 October 2010 (UTC)

I think that the jobs section may benifit from a inner-wikipedia link to the Griggs v. Duke Power Co. were the supreme court found that under Title VII of the Civil Rights Act, if such IQ tests disparately impact ethnic minority groups, businesses must demonstrate that such tests are "reasonably related" to the job for which the test is required. Basically you can not use an IQ test to determine job placement. 71.201.50.30 (talk) 03:30, 3 January 2011 (UTC)Ben

Clarification requested

Hi, could someone clarify this sentence, which appears in the section "Genetic influences": Heritability measures the proportion of variation that can be attributed to genes within any measured population (however defined), and not the extent that genes contribute to intelligence. I think this sentence is trying to make an important distinction, but I'm not sure what the distinction is -- to me it sounds like "proportion ... attributed to genes" is the same thing as "extent that genes contribute to intelligence." Or is "attributed to" intended to refer to correlation while "contribute to" is intended to imply causation? Duoduoduo (talk) 15:58, 20 December 2010 (UTC) ..... I see that the article Heritability says in its intro: Pay close attention to the variation part of "phenotypic variation": if a trait has a heritability of 0.5, it means that the phenotypic variation is 50% due to genetic variation. It does not imply that the trait is 50% caused by genetics. Again, I'm confused: if "50% caused by genetics" does not mean the variation is 50% due to genetic variation, what else could it mean? My memory of my past reading on this topic makes me think that this passage has some debatable POV in it. Or does "50% caused by genetics" have some specific and different meaning that is widely accepted? Duoduoduo (talk) 16:38, 20 December 2010 (UTC)

I'm not sure what the formulation means. The author has probably tried to communicate the idea that heritability does not tell us how much genes contribute to the phenotype in any particular individual. I modified it accordingly.--Victor Chmara (talk) 17:18, 20 December 2010 (UTC)

Okay, now I think I know what it means. Suppose there's a population of human clones with identical genomes. The variation in IQ between these clones is entirely due to environmental factors, i.e. the heritability of IQ is zero. This, however, does not mean that genes do not contribute to IQ in these individuals. Similarly, if it was possible for people live in an identical environment, all of their IQ differences would be due to genes (100% heritability), but the environment would still influence their IQs (equally for all). Heritablity indicates how much genes and environment contribute to variation in a trait, not how much they contribute to the trait.

This could probably be described in an easier to understand way in both articles.---Victor Chmara (talk) 17:43, 20 December 2010 (UTC)

I see. How about "Heritability measures the proportion of variation that can be attributed to genes within any measured population (however defined), and not the extent that genes contribute to intelligence. For example, if both genes and environment have the potential to influence intelligence, but if a given sample of individuals shows very little genetic variation and a great deal of environmental variation, then the contribution of genetic variability to phenotype variability in that sample will be lower than if the sample showed greater genetic variability."? Duoduoduo (talk) 20:15, 20 December 2010 (UTC)

Mental disorders & how they relate to IQ

Such as:

-ADD and ADHD -Autism —Preceding unsigned comment added by 74.171.79.234 (talk) 05:29, 23 December 2010 (UTC)

The proliferation of "IQ and environment" articles

There are way too many overlapping or otherwise unsatisfactory articles about environmental influences on IQ. Firstly, there's Environment and intelligence, which has lots of material similar to the much better written Heritability of IQ. Then there's Malleable intelligence, which is a collection of mainly neurobiological findings that may or may not be related to "malleable intelligence". Finally, there's IQ testing environmental variances, which is a crude POV fork of Race and intelligence.

I suggest we do the following:

• Delete IQ testing environmental variances entirely.
• Merge Environment and intelligence with Heritability of IQ. The article could be renamed to Genetic and environmental influences on intelligence or perhaps Nature and nurture of intelligence.
• Merge the worthy bits of Malleable intelligence with Neuroscience and intelligence or some other article(s).

--Victor Chmara (talk) 19:49, 19 January 2011 (UTC)

That "IQ testing environmental variances" article is pretty awful. Maybe someone should nominate it for deletion. Your other suggestions sound good too. There seems to be little logic to how information is divided between these articles. Perhaps it would be helpful to create a Wikipedia project on intelligence to help prevent info duplication among articles?-SightWatcher (talk) 04:24, 20 January 2011 (UTC)

I just proposed IQ testing environmental variances for deletion.--Victor Chmara (talk) 15:38, 25 January 2011 (UTC)

I agree that the article is horrible - but that is not a reason for deletion. The topic is clearly notable and there is a wide lterature about it. It could be a merge is in order - but I think that there will eventually be room for an article that describes only the many, many ways in which intelligence correlates with enviromental factors - we have one on Heritability of IQ - this one seems like its logical twin. ·Maunus·ƛ· 16:55, 25 January 2011 (UTC)

I've had my eye on some of those articles for a while as candidates for merger with Heritability of IQ, which has long had its own set of problems. At the very least, there should be redirects from previous article titles to appropriate sections of a newly sourced and improved article on what "heritability" means and what reliable secondary sources on the issue of heritability of IQ say. P.S. There is already a WikiProject Psychology that needs more participation by more editors, and especially more participation by more editors who are familiar with the secondary literature on psychology. -- WeijiBaikeBianji (talk, how I edit) 17:58, 25 January 2011 (UTC)

Maunus, the article is not about environmental influences on intelligence in general. It is specifically and only about environmental influences on racial differences in intelligence. Heritability of IQ deals with genetic and environmental influences (mainly within populations), and there are two other articles, linked to above, that deal with environmental influences. IQ testing environmental variances is not a "logical twin" of the heritability article, but rather a bastard child of Race and intelligence from which it was originally cut and pasted. There is nothing worthwhile in it that is not already included in other articles.--Victor Chmara (talk) 18:15, 25 January 2011 (UTC)

No, not as it is written now. But the question is whether the topic deserves an article - I think it does and it already has one and that is Environment and intelligence. I suggest that the sources and information included here, such as the material from Jencks, Flynn and other sources info be merged to that article.·Maunus·ƛ· 23:34, 25 January 2011 (UTC)

There's an ongoing deletion discussion concerning the environmental variances article. More people should take a look at it.--Victor Chmara (talk) 13:02, 7 February 2011 (UTC)

Miradre's edits to the history section

I disagree with edits such as this[5][6][7]. Generally it seems that Miradre is removing every connection between scientific racism and the early history of IQ measurement - this amounts to falsifying history as there was an intimate connection between the development of IQ measurement and racialist science and eugenics. The argument that we have another argument about race and intelligence does not mean that it should nt be mentioned here. It is part of history of IQ (even psychology textbook's mention this part of it) therefore it should be covered here. I am of a mind to revert Miradre's latest 20 edits or so, but post here instead to let him rather show good faith and do so himself. ·Maunus·ƛ· 12:16, 17 March 2011 (UTC)

If you prefer we can give a link to the History of the race and intelligence controversy.Miradre (talk) 12:28, 17 March 2011 (UTC)

No, I'd prefer to give a historical treatment of the development of IQ measurement that doesn't attempt to hide the fact that it started out as a less than noble venture. Also you have turned rather coherent prose into an incoherent list of events. Its not a bullted list but it reads like one. This isn't good writing, a history section should be developed as a cohesive historical narrative.·Maunus·ƛ· 12:34, 17 March 2011 (UTC)

Previously there was no mention of most of the history of the IQ tests. There were several factual errors as documented in my edit summaries. Obviously we cannot duplicate the whole History of the race and intelligence controversy here but I will add a link.Miradre (talk) 12:39, 17 March 2011 (UTC)

I will provide the needed historical context of the events described.·Maunus·ƛ· 12:45, 17 March 2011 (UTC) (UTC) (UTC) (UTC)

Link to a vaild online IQ test

Is there an online site, such as an university, that offers one or more of the major free IQ test that is not marketing based. I know that it would greatly enhance the article if the readers were able to reliable test their own IQ score. I don't know how the Wikipedia rules work for suggesting an outside souce that is accurate, but I think universities would be the first place to start.

These are the reliable tests that I was thinking about: Wechsler Adult Intelligence Scale, Wechsler Intelligence Scale for Children, Stanford-Binet, Woodcock-Johnson III Tests of Cognitive Abilities, or the Kaufman Assessment Battery for Children-II, 71.201.50.30 (talk) 23:31, 2 January 2011 (UTC)Ben

Wikipedia isn't in the business of advertising sites, just in providing an encyclopaedia with information about a subject. A test yourself site does not provide any extra encyclopaedia type information. Dmcq (talk) 23:51, 2 January 2011 (UTC)

Ok that i understand, I think then there needs to be focus on the popular culture usage section of the article. It states that there is a lot of poorly constructed online IQ test but does not offer information about good tests online or what makes a test online valid. Are there good test online? Can your IQ be accuratally measured by a test online? 71.201.50.30 (talk) 03:13, 3 January 2011 (UTC)Ben

Are there WP:Reliable sources that discuss that point? Find some citations first. Dmcq (talk) 11:35, 3 January 2011 (UTC)

I've added a bit of welcome data to User talk:71.201.50.30 since it looks like you might like to edit articles. You don't have to read it all, I never did, but the principles in WP:5P are always worth a read. The applicable one here is pillar 2, citing sources. Dmcq (talk) 12:02, 3 January 2011 (UTC)

My understanding has always been that there is no such thing as a valid online IQ test. It should be administered by a trained professional. It would be no more legitimate than a Cosmo 'does your guy have the hots for you' quiz. Bakkster Man (talk) 13:36, 13 April 2011 (UTC)

???

if you think that the history section should be edited, then why don't you edit it?

---

— Preceding unsigned comment added by A.a.p.cool (talk • contribs) 00:41, 25 May 2011 (UTC)

Unprofessional and possibly racist remark I can't remove

Section 14.2 on Race has the last sentence looking like this for me:

"In contrast, other researchers such as Richard Nisbett argues that environmental factors can explain all of the average group differences. (wild stab in the dark .... which race has the lowest iq ?)"

I think the parenthetical statement is wholly unnecessary in this section, but I can't see it on the edit page for some reason. Perhaps I don't have the clearance to edit this? Someone please fix this though; it's a pretty offensive remark on a controversial topic. SnehaNar (talk) 02:12, 26 May 2011 (UTC)

It's already been removed by ClueBot with this edit. Shearonink (talk) 12:16, 26 May 2011 (UTC)

New data

I added a short paragraph in the real-life accomplishments section describing the data provided by a survey of the members of Mensa Finland, because it seemed to contradict the existing data significantly. The actual report is not available publicly on the web, but I think anyone interested can contact the Mensa Finland office by e-mail: toimisto at mensa dot fi, and ask if they can send it to you or refer you to the authors. Subarctica (talk) 22:52, 8 July 2011 (UTC)

Mensa members are self-selected, and the vast majority of people with high IQs never apply to Mensa. The members are not representative of the high-IQ population. If the numbers you provided are from a survey of Mensa members, how come the mean IQs are so low? Finally, I don't think your source, an unpublished internal survey, can be regarded as reliable.--Victor Chmara (talk) 23:21, 8 July 2011 (UTC)

You are absolutely correct about the self-selection, but since this is actual raw data (first I've seen anywhere) instead of a subjective assessment, I thought it would be very noteworthy. About the numbers, you must be misunderstanding something. The numbers shown describe the whole population, and are calculated via statistics based on the answers. I will consider clarifying the section so that there is no possibility of misunderstanding. The report was published in the official members-only journal, in which the authors do not have any editorial power. I want to emphasize that this is raw data, not data interpretation, and therefore it should not be a big problem that the journal is not peer-reviewed. Moreover, the journal does not promote any specific point of view. I would compare it to any other official interest group journal, and in my opinion it is therefore reliable. Subarctica (talk) 23:52, 8 July 2011 (UTC)

Actual "raw data" from where? From everybody who has taken a Mensa test? How were the representativeness of samples, statistical significance, etc. assured? I think a mean IQ of 110 for college graduates (I assume this means maisteri or equivalent) is implausibly low, as is, perhaps, the 117 for PhDs/MDs. In the Finnish WAIS-III standardization sample the mean IQ for high school graduates (ylioppilas) was also 110, and American data suggest a mean IQ of as much as 115 for college graduates, IIRC. What is the norm sample Mensa Finland uses? Please read the Wikipedia policy on reliable sources, I don't think any of your arguments suggests that your source is reliable.--Victor Chmara (talk) 00:23, 9 July 2011 (UTC)

The answers comprised 40-50 % (can't find the exact number now) of Finnish members. I'm not here to explain or argue the methodology, which is thoroughly explained in the study itself. The college-equivalent classification was mine, consisting of AMK and university degrees. Personally I don't think 110 is "implausibly low", given that it is the 75th percentile. Approximately half of the Finnish population goes to academic high school (lukio), and although I don't have the numbers here, the part of them who continue to AMK/university should be of the order of 1/2. Therefore, the 75th percentile is entirely logical, even if we correctly assume the high school or AMK/university selection do not select high IQ perfectly.

However, what is important here is that it does not matter what I or you think is a low or a high number. The only options are to believe the data, or challenge its credibility openly, hopefully by contacting the authors so that they can defend themselves. I couldn't find the WAIS-III source you mentioned, but in all the American sources I have looked at, the classification seems to be done by assuming SAT percentiles correspond to IQ percentiles, in which case it is obvious that college students will get high values.

I have read the page on reliable sources, and I still have to disagree with you. The Mensa Finland journal is a high-quality publication by a well-known society. It is not a peer-reviewed or a research journal, but that does not automatically mean it's not reliable. Perhaps you could help by providing specific arguments as to why would it not be reliable?

Subarctica (talk) 01:08, 9 July 2011 (UTC)

In fact, the college graduates part of the US data presented in the article can easily be seen to be mathematically impossible. Claiming the mean IQ is 115 means the mean college graduate is in the highest 1/6th of the population. However, in the US, approximately 63 % of the age group goes to college, and 56 % of them graduate (2009 US educational attainment gives 30% of the population over 25 years of age having at least bachelor's degree). That, assuming symmetrical distribution and any reasonably well-behaving correlation between IQ and educational, is equivalent to saying that 1/3 of the population belong to the top 1/6, which is simply not possible. It's therefore not clear how such number has even been permitted to be published. Since WP:NOR allows routine calculations (which, I hope, can be understood to include trivial statistics), I would suggest mentioning that the number given is likely to be not correct. I would like to hear opinions if this would be okay. Subarctica (talk) 12:17, 9 July 2011 (UTC)

All right, Wikipedia:These_are_not_original_research tells that statistics generally are not routine calculations. Therefore, I partially retract my proposition that the 115 number is stated to be inaccurate. However, the same page later states that the routine level is determined by the nature of the article in question. And since IQ is defined by statistics, I would assume that anyone with serious interest of the subject (IQ, not intelligence in general) would consider simple statistics routine. However, this needs more opinions than mine. Subarctica (talk) 12:28, 9 July 2011 (UTC)

Well, I don't think routine calculations should be left to you. If the mean IQ for college grads is 115, and a third of the population has a bachelor's degree or higher, it does not mean that a third of the population has IQs in the ≥115 range; half of them have IQs <115, assuming normality.

In "IQ Testing 101", Alan Kaufman reports the mean IQ for college grads in the US as 115, and as 125 for PhDs and MDs. These numbers are based on data from Wechsler standardizations.

Where was the Mensa study published? Please provide name of journal, issue number, and page number. I don't have access to the study, so could you please tell me how they decided that average IQs for college grads and PhDs/MDs are 110 and 117, respectively, if they only had data on people with IQs in excess of 130? I'm genuinely interested. If your number for college grads is for both university and AMK grads, it's plausible, but it's also original research if it's not reported as such in the study.--Victor Chmara (talk) 15:38, 9 July 2011 (UTC)

It does seem that I made a silly mistake of thought. You are of course correct, that if the mean is 115 (and the distribution is normal), only half of the group actually are above that. Therefore, the 115 would not be impossible after all, but only if the college admissions/graduation reliably selected high IQ. This is a strong assumption, and does not apply in Finland: 45% of the members who answered do not have post-high school degree, even though 92% of them were over 25 years of age.

WAIS seems to test aspects such as vocabulary, general knowledge and arithmetic in addition to logical reasoning, and thus it measures a somewhat different thing than Cattell Culture Fair III test that was used for a long time in Mensa Finland. This further explains the score of 115; if the logical reasoning IQ is combined with variables that correlate more strongly with success in school (such as vocabulary, general knowledge and arithmetic), educated people are likely to achieve scores greater than their pure logical reasoning IQ, which would be 110 based on the MF survey. In addition, WAIS standardization sample sizes were extremely small, approximately 7*10^(-6) of the US population for WAIS IV.

The MF study was published in Mensalainen - National Newsletter of Mensa Finland (published by Mensa Finland, ISSN 14553422), spread to issues from 4/2008 to the later half of 2010 (I couldn't find all the issues). The "Comparison with the general population" part, in which the education was included, of the study was published in the issue 2/2010, pages 22-29. The population averages are calculated with conditional probability - find the probability that a person in the general population having college degree has IQ >= 130, calculate the percentile corresponding to IQ 130 from that, and finally assuming the normal distribution, find the mean IQ. It is not as good as a real mass-measurement would be, but in my opinion the methodology was straightforward and plausible. In any case, at least when considering the logical reasoning IQ, I would consider the MF study much more reliable than WAIS with its wider focus and extremely small normalization sample.

I disagree with the composite number being original research, although it may be close to the borderline. NOR permits compiling facts as long as that is not done in a manner implying anything the sources do not imply. I think that combining AMK and universities is the only sensible way to achieve the equivalence to the US college, and that doing so is not implying anything. Of course, it is possible (and in fact shown by the MF study) that the people with university degrees have higher IQ than the combined university-AMK pool, but as most of the readers do not know the difference, I originally saw no reason to include that statement in the short paragraph. I'm not sure if the section should be expanded, but the data provides much additional comparison of the members to the general population, which might be helpful for the article.

Subarctica (talk) 17:51, 9 July 2011 (UTC)

Fluid and crystallized intelligence measures are more or less perfectly correlated in culturally homogeneous populations such as Finns, so I don't think the test used (WAIS vs. CFIT) makes any difference.

The WAIS-IV standardization had a sample size of 2,200, and Kaufman used several samples. That's a lot more people than in the Mensa study (Mensa Finland has about 2,500 members and the response rate was 50 percent or less). What's more important is that standardization samples are nationally representative, whereas the Mensa sample is, at best, representative only of Mensa members. In The Bell Curve Murray and Herrnstein used the NLSY sample (n>11,000), and found that the mean IQ for college grads was about 113 (graph on page 46).

Murray and Herrnstein also showed that about 70 percent of people in the top IQ decile are college grads; the percentage is even higher for the top two percentiles (pp. 36-37). Almost 40 percent of all Finns aged between 25 and 35 have a tertiary education degree. The fact that 45 percent of Mensa respondents do not have one suggests that there is truth to the stereotype that high-IQ societies attract people whose only accomplishment in life is a high IQ. To the extent that Mensa members are academic underachievers compared to non-members with similar IQs, the method used in the study will underestimate the average IQs of college grads, PhDs, and MDs.

For the following reasons, information from the Mensa survey should not be included in the article:

• Wikipedia articles should be based mainly on reliable secondary sources. The use of primary sources is discouraged.

• It would be particularly unwise to use a primary source like the Mensa survey which was published in a non-academic, non-refereed journal.

• The methodology of the Mensa survey is dubious and its results contradict those in reliable academic sources, as detailed above.

I removed the paragraph from the article.--Victor Chmara (talk) 11:26, 10 July 2011 (UTC)

I guess this must all depend on the percentage of people who go to university and get degrees. It used to be far lower in the past. I hadn't heard that stereotype about high-IQ societies, I can see it might be a way of getting company of comparable intelligence if it isn't available otherwise. 14:08, 10 July 2011 (UTC)

Yes, age is a big confounder here, too. I don't know how they dealt with it in the Mensa survey. As to who would join Mensa, I think it goes without saying that they are not a random sample of high-IQ individuals.--Victor Chmara (talk) 15:25, 10 July 2011 (UTC)

Are you implying that in nonhomogeneous populations the fluid intelligence scores would be less than the crystallized intelligence scores for highly educated people? I tried to find sources for your claim, but couldn't. If there are some, wouldn't you think that would bear mentioning in the article.

The age-education distribution was not shown in the journal, but since the average member age is somewhere near 40s, the age might explain some of the education results

The education part of the MF survey had a sample size of 1077, which is 1 % of Finnish high-IQ population. The NLSY survey seemed to have not concentrated on the IQ: only 1 in approximately 5 to 10, depending on the overlap, had taken any IQ test, and while browsing through the data table I noticed there were extreme discrepancies between scores. For example, ID 7452 had achieved 43rd percentile in one test and 88th in another. I will try to get hold of the book to see how they calculated the individuals with differing percentiles or assessed the reliability of the various tests. As to the absolute sample sizes, I am under the impression that the percentage of population is what matters more. Therefore, the 7 parts per million of WAIS-IV or the similar number for NLSY would not be in effect a lot more than the 1 %.

I hope that you keep your personal opinions about the reasons people join high-IQ socities to yourself in this kind of a discussion. The fact that I or Dmcq above haven't heard of it hints that it's not a very wide-spread stereotype. However, I agree that if there was data showing that Mensa members are underachievers compared to the general 98th percentile population, the effect would be as you described. However, since I've never heard of any kind of evidence supporting that claim, I find it hardly plausible. I do agree that it is very possible that Mensa is not a random sample of high-IQ individuals. However, goes without saying is not a very scientific or encyclopediatic claim, and until there is any kind of data about it, the factual state of it must be determined unknown.

I don't agree with calling popular science books academic sources, but I (still) agree that the Mensa journal was a borderline case for source reliability. However, specifically since it contradicts data from other sources, I believe the data should be mentioned as alternative results, to provide a more neutral view instead of just the results of one study. However, I will wait for more opinions before re-adding the paragraph.

Subarctica (talk) 11:39, 11 July 2011 (UTC)

Please note that I added numbers based on two different IQ tests (Kaufman Adolescent and Adult Intelligence Test and Kaufman Brief Intelligence Test). Both are reported very clearly in Kaufman & Lichtenberger's book (pages 121 and 122), so there should be no problem. I added them before the WAIS scores due to them being much more recent (it seems that the data used in that section originally comes from Reynolds et al. unpublished study in 1987). The book doesn't seem to report the PhD/MD IQ for KAIT/K-BIT. Subarctica (talk) 16:20, 11 July 2011 (UTC)

If Cattell's theory is correct, the connection between fluid (Gf) and crystallized (Gc) intelligence is that you "invest" your Gf in acquiring Gc, or knowledge. The more homogeneous the opportunities to learning are in a culture, the more highly Gf and Gc will be correlated with one another. In any case, all IQ tests are primarily measures of g, which means that all IQ tests are highly correlated.

More than 11,000 people had their IQs measured in the NLSY survey, and those results have been and continue to be used in numerous studies. The test used was the AFQT, one of the best validated tests ever. No IQ test has a test-retest reliability of 1, so of course there's going to be some fluctuation. The AFQT has a test-retest reliability of 0.9.

The Mensa sample had zero people with IQs in the typical range for college graduates, so obviously standardization samples that have plenty of such people are going to be much more reliable sources for information on their IQs.

On the Mensa member stereotype, here's the New York Times[8]:

Mensans bristle at the prevailing stereotype about them: that they are lonely losers who rarely wash and who exhibit all the social skills of Dustin Hoffman's idiot savant in "Rain Man."

In any case, you cannot expect that a self-selected sample will automatically be representative of the relevant population. My conclusions are not dependent on "popular science books", all are backed up by scholarly sources.--Victor Chmara (talk) 16:07, 16 July 2011 (UTC)

Cattell's theory sounds very logical, and I believe you are more of an expert of the subject than I am. I don't believe the homogeneouity (sp?) assumption is necessarily true even in Finland due to geographical differences in education quality/specialization. In addition, the fact that one can train for Gc tests much better than for Gf tests results - according to my logic - Gc tests measuring motivation and dedication in addition to Gf. Moreover, would you (or authors of the field) agree that Gc correlates with general educational success when Gf is accounted for? That is, if two person of equal Gf have differing results in school, the one with better results probably also has higher Gc?

The mistake about the small sample size was mine - NLSY 79 did not list AFQT as an IQ test in the NLS Investigator. I'm not sure how much to read into that, but at least the current AFQT seems to measure only Gc (2/3 of the points come from paragraph comprehension and word knowledge). I wouldn't think the examinees in the NSLY specifically trained for the test, though.

To me it seems that the NYT stereotype does not tell about the underachieving, because the word loser in social context usually means someone with no friends and so on, with no relevance to job or education status. I also think that the rest of the NYT article could be used as counter-evidence to your original stereotypical claim.

I hope I haven't sounded like I am strongly assuming Mensa members represent the total top 2 % population well. My point has been that until there is any kind of information about how the members are compared to the non-members, the uniform assumption might be the safest one (i.e., the expected error would be the smallest compared to any other assumption). However, when using such an assumption, it should be remembered (and mentioned) that it is a weak one. For example, something like this:

Were the Mensa members good representatives of the total 98th percentile population, the average Gf IQ of a college graduate...

I am not claiming the other information should be removed, as is obvious from my original edit. However, I understand your arguments for not including this information, given that you seem to believe that the Mensa population is very significantly different from the general top 2 %. Since there is no consensus, I believe the original state of the matter (i.e. no Mensa survey information included) should be preserved for the time being.

Subarctica (talk) 14:59, 17 July 2011 (UTC)

Replacing Intelligence_quotient#Sex with summary of main article

Currently the content of Intelligence_quotient#Sex appears to contradict the content of the main article Sex and intelligence. I edited the section according to summary style, removing the content based on primary sources. This has been reverted with what looks like an incomplete edit summary. Discuss? aprock (talk) 23:23, 26 July 2011 (UTC)

Given that the lede to Sex and psychology was changed after the revert, I've redone the summary. If there is a problem with the summary, I encourage constructive editing in lieu of reverting. The prior summary is a mess of misused primary sourcing and not a summary of the main article at all. aprock (talk) 23:46, 26 July 2011 (UTC)

You put in something that was further away from reality than what was there before. Why should I keep what you wrote and try and develop it? You did the bold in WP:BRD. I did the revert. This is the discussion. You can always retrieve things from he history. I'll copy over that bit about maths and language to the sex and psychology lead, I'd been thinking it was missing something like that. The American Psychological Association's 'Intelligence: Knowns and Unknowns' can be used as a general citation for everything in the section. Dmcq (talk) 09:32, 27 July 2011 (UTC)

The "something" that I put in there was summarized from the main article. By all means fix the content in both the main article and the section. I see that you've done a bit of that already, and I appreciate the constructive help. Thanks. aprock (talk) 14:30, 27 July 2011 (UTC)

Health and intelligence

Just checking, did you actually read what Health and intelligence referenced? It references a disambiguation page and I'm rather concerned that a large section has been removed without reference to the article whichever of them is referenced. Was there extra information here and if so has it just been deleted? Dmcq (talk) 19:04, 1 August 2011 (UTC)

Without being sure, I'm going to guess that you're referring to this edit: [9]. If that's the case, I replaced a non-summary section with summary of the article. I did the original summary based on the first article of the redirect and forgot about the second. I've fixed the redirect, added a link to the second main article, and expanded the summary. If you feel that specific content was removed that shouldn't have been, by all means add it back to the appropriate main article. As best as I can tell, all the important stuff is in the main articles. If you feel the summaries are incomplete in some way, by all means expand them. aprock (talk) 20:40, 1 August 2011 (UTC)

Did you check the content which you were replacing or did you replace it without checking? It is a bad idea to just discard peoples work without checking it first for good content which isn't in the stuff replacing it. You replaced it with reference to another article, dif you check the referenced article contained the content? If not I'll just revert it. Dmcq (talk) 21:05, 1 August 2011 (UTC)

Yes, I checked. aprock (talk) 21:43, 1 August 2011 (UTC)

Thanks. Dmcq (talk) 22:05, 1 August 2011 (UTC)

Ordinal Data and Statisticis

I went ahead and added this section in criticisms to point out that with ordinally scaled data it is classically innapropriate to deal with means and standard deviations. It's a tricky subject because it applies to most psychometry out there, and there's an ongoing discussion about treating it as interval data (although in general a lot of people just ignore the concept of levels of measurement and do whatever they feel like). Theres a pretty solid reference linked for it that talks about the conversation.

I wasn't sure if this deserved to be in the criticism part of it or somewhere else, but it's a pretty important point. Any time you start doing research like looking at differences in averages and stdevs across demographics you start crossing the line of what is and isn't appropriate to do with ordinal data.

Anyways, I added it once and to be fair the starting edit I put in was pretty lousy with no references (beyond a link to ordinal scales), but I believe the current reference should suffice. — Preceding unsigned comment added by 98.201.100.117 (talk) 01:12, 10 September 2011 (UTC)

Moved topic to end.

I have removed the section as there is no mention of intelligence or IQ in the citation. You are quite right in principle but it really makes no discernible difference in this case. By the way if you want to get into the actual statistics rather than start from that paper the terms to start with are non-parametric statistics and latent variables. Where a lot of small factors contribute to something then the result tends in most cases to a normal distribution anyway and that seems to be pretty much the case for intelligence. In actual fact a log-normal distribution might be a better model in this case but it wouldn't make much difference as IQ is only use in an ordinal sense anyway. It may be that there is an underlying scale which is closer to the square root of the IQ value rather like skin area depends most on height, but that is really irrelevant to the uses to which IQ is put.

If you can find a citation which actually mentions this as a problem with IQ or intelligence measures rather than psychometry in general then please add the section back, otherwise the article Psychometrics is the appropriate one I believe. I'll stick a 'see also psychometrics' in the 'Outdated methodology' section so anything there can be found. Dmcq (talk) 11:14, 10 September 2011 (UTC)

The normality isn't really relevant since it is assumed by definition, in fact it's impossible to know since it's normal because it's designed as normal (they use the percentiles and the stdev of a normal curve to define score). Anyways, just a tangent. I see your point about this being a broader issue with psychometrics as a whole instead of IQ in particular. It just bothers me that the issue isn't raised in some way. As a side note, considering that normality is assumed by definition (the resulting scores then, by design, should be normally distributed) this is an interesting article (although it is an older one and the scales have probably been refactored) http://www.abelard.org/burt/burt-ie.asp original paper starts like 1/2 way down.

What bothers me the most is that there are these assumptions that the differences found in the metric show conclusive evidence and even worse, causal relationships (god forgive the scientist who does this), when I think there is still significant arguments out there to question the validity of the metric itself.

By v0v thats well beyond the scope of the purpose of this discussion section. — Preceding unsigned comment added by 98.201.100.117 (talk) 02:38, 12 September 2011 (UTC)

I do hope you realize Cyril Burt died 40 years ago and his main work was long before that. IQ test raw scores are shoehorned into a normal curve nowadays by looking up an associated table. Personally I feel he tried reading far too much into those scores. More can be extracted from the raw test results I'm sure but that's really for the psychometrics article to go on about. Dmcq (talk) 18:23, 12 September 2011 (UTC)

Perhaps the best way to approach this is within the history section. IIRC, IQ was originally intended to identify under-pwerforming students for remedial help. In such a case, it was beneficial to identify the bottom x% of students. Perhaps this makes for a more rational location to discuss why the ordinality is the way it is, why normal distribution is assumed, and why the precise spread between two individuals IQs does not have a quantitative meaning. Bakkster Man (talk) 17:14, 12 September 2011 (UTC)

Also, a better source for this argument would be The Mismeasure of Man, as this general topic is covered in that book. aprock (talk) 17:25, 12 September 2011 (UTC)

But see that's what bothers me about the whole criticism section, it's composed of a various conversations taking place in pop-science books as opposed to scholarly articles discussing more meat and potatos stuff. I like and respect the dudes that wrote those books, but they shouldn't be the primary references for this article if there is long term interest in being anything more than another pop-sci article. TLDR? There's a reason academics don't cite those books. — Preceding unsigned comment added by 158.35.225.240 (talk) 22:55, 12 September 2011 (UTC)

Welcome to the world of secondary sources. aprock (talk) 00:27, 13 September 2011 (UTC)

Incorrect use of statistical term "median" corrected with correct word mean (or average)

In statistics there are three primary measures of central tendency: mean, mode, and median. These terms are frequently and erroneously interchanged by popular media, but in a scholarly article should never be. In this article the proper statistical term is mean (which can also, legitimately, be called "average"). I would ask the author(s) of this article to do some checking about standard normal distributions before reverting this important edit. — Preceding unsigned comment added by N0w8st8s (talk • contribs) 18:02, 20 September 2011 (UTC)

Unbalanced

The article is unbalanced in its presentation of criticism against G. There are many serious intrelligence researchers from Sternberg to Gardner who dispute the usefulness of G as a measure of general intelligence. They should be included. The very short criticism section presents the criticism as if Gould's is the most important and the presents a length quote from Flynn as the last word, giving the distinct impression that any critique is unfounded. Psychology textbooks do not generally present a view of IQ as an unproblematic measure of cognitive ability, neither should this article.·ʍaunus·snunɐw· 12:18, 10 September 2011 (UTC)

There's an article General intelligence factor. Shouldn't that be the appropriate article with just a short summary here with a see also at the top? Also this article is about IQ rather than g so I'm not certain why the section is there at all except to note it is a general argument against measures of intelligence. AFter all g and IQ are different measures of intelligence. Dmcq (talk) 12:25, 10 September 2011 (UTC)

Sternberg does not dispute g. This is what he wrote in his article in The Scientific Study of General Intelligence: Tribute to Arthur R. Jensen (2003):

The evidence in favor of a general factor of intelligence is, in one sense, overwhelming. This evidence is so well documented by Jensen that there is no need to repeat it here. One would have to be blind or intransigent not to give this evidence its due. Not only is there evidence for the internal validity of the g factor, there also is evidence for its external validity as well. Again, Jensen's documentation, as well as that of others, is scientifically impressive. The impact of Jensen's work on g to the field of psychology—in terms of both the support and the criticism it has generated—is a tribute both to Jensen and to his many ideas, including that of a general factor.

Sternberg's shtick is to claim that there are types of intelligence that are not correlated with g. He does not claim that g does not exist. He says that besides analytical intelligence (=g), there are creative and practical intelligence. However, I don't think anyone besides Sternberg and some of his students actually believe in this triarchic theory because there is very little empirical evidence behind it.

Gardner, in his turn, is not an intellectually serious researcher. He is hostile to all attempts at empirically investigating his myriad intelligences, and while his work has influenced some educationalists, he is a non-entity in academic psychology.--Victor Chmara (talk) 12:57, 10 September 2011 (UTC)

Sternberg disputes the "g-ocentric" view of intelligence as he has published in several different papers he thinks that G is not the whole story of cognitive abilities. I did not say that he disputes its existence. Whether you think that "nobody" believes his thriarchic theory is utterly irrelevant - his theories and those of Gardner are presented in most psychology textbooks treatment of Intelligence. Your arrogant characterization of Gardner is also not relevant or helpful - [ and it seems rather arrogant to me ]- as he is also frequently cited in psychology textbooks.^{this comment has been redacted for enhanced civilty} ·ʍaunus·snunɐw· 13:10, 10 September 2011 (UTC)

Please try to avoid characterizing other peoples comments as 'arrogant' or suchlike, see WP:CIVILITY. Dmcq (talk) 15:18, 10 September 2011 (UTC)

I did not characterize Victor as arrogant, I characterised his statement as such, which it clearly was. Please try to avoid giving bogus civilty warnings. Referring to professional scientists as having a "schtick" or as being "a non-entity" is arrogant, does not help build consensus.·ʍaunus·snunɐw· 15:20, 10 September 2011 (UTC)

Speaking as someone who actually understands the mathematics of what's going on, g is not a physical thing, it is a model parameter. While g may (or may not) be an effective model of intelligence, to suggest that it is a physical thing is nonsense. Some might discuss whether or not that model parameter correlates with something that is physically real, but g is not a physically real thing. I am honestly surprised that this sort of implication is even in the article. aprock (talk) 17:33, 12 September 2011 (UTC)

High standardized test scores are best viewed as a sufficient but not necessary condition to determine how well someone's mind works. For instance, I know many smart people who test poorly and many people who scored very highly (nearly perfect) on tests like the SAT and GRE and also highly on iq tests. They were also very intelligent. Smart people can do poorly on these tests. Dumb people cannot ace them. I believe this bit of common sense is lost in the debate over correlation with g , etc. I believe most of the correlation breaks down when smart people do not test well. Wvguy8258 (talk) 19:47, 5 November 2011 (UTC)

Neuroscience/Neuropsychology resource

As Brain Changes, So Can IQ; Study Finds Teens' Intellects May Be More Malleable Than Previously Thought OCTOBER 20, 2011 by ROBERT LEE HOTZ ... new findings by researchers at University College London, reported online in Nature ... 97.87.29.188 (talk) 23:56, 20 October 2011 (UTC)

table showing different iq scores per individual

Does this table make any sense given that various test have different standard deviations and perhaps even centers? It would be more useful to report different iqs in terms of percentile. For instance, a stanford-binet iq score will usually be higher than a wechsler due to different variation. Wvguy8258 (talk) 19:41, 5 November 2011 (UTC)

IQ tests are standardized to have the same distribution nowadays. And even when they weren't the difference was something like a standard deviation of 13 or 15. Even that would make a trivial difference to the overall tenor of those results. Dmcq (talk) 20:06, 5 November 2011 (UTC)

(Some versions of) Cattell Culture Fair uses SD 24, a significant difference over 15/16. I'm not sure if they are used that much anymore, but certainly historically there is a large variance of different SD's. Subarctica (talk) 22:56, 5 November 2011 (UTC)

Race

An editor is putting a table into the race section which doesn't summarize anything in the race and intelligence article which is the main article about that. This article should not have a whole load about a subtopic which is not in the main article about the subtopic. In fact it should not have a whole load about the subtopic in the first place and especially not tables, it should just summarize what the subtopic says - the relevant points in the subtopics lead. Dmcq (talk) 14:37, 5 December 2011 (UTC)

Well I've reverted it twice but the editor seems to think I'm on some ideological mission. Amazing the way people attribute to others what's most probably true of themselves. So over to somebody else to assess. Dmcq (talk) 14:55, 5 December 2011 (UTC)

Hello

How accurate are IQ online tests? Thanks — Preceding unsigned comment added by 213.198.207.155 (talk) 03:25, 12 December 2011 (UTC)

Stanford or Emory?

Next to Lewis Terman's name there's a redlink - [[Emory University School of Education|Stanford University]] . Emory University is in Georgia, there's no mention of Emory on Terman's page but there is Stanford_University_School_of_Education - is that what should be linked here? --24.23.193.132 (talk) 11:58, 28 December 2011 (UTC)

Brain size and weight

There is a small but significant link between brain weiht and intelligence, see Height and intelligence. The following is wrong plus anyway a citation would be needed for the following which was inserted:

"Brain size is not correlated to intelligence: Anatole France (1017 gr), Einstein (1209 gr), Gauss (1490 gr), etc... being the main value 1300-1400gr. (It seems someone erase this sentence, but it is true, so i'll continue to put it here). People that correlates brain size and intelligence are doing pseudoscience, called: phrenology. Specifically: http://en.wikipedia.org/wiki/Phrenology#Method and http://en.wikipedia.org/wiki/Phrenology#Pseudoscience"

The ip has been warned about the threat to edit war and directed here. Dmcq (talk) 20:15, 6 January 2012 (UTC)

g and Cattell-Horn-Carroll theory not modern theories?

They are the mainstream theories in IQ research and Cattell-Horn-Carroll theory is a new as any of the others mentioned. I propose correcting this. Acadēmica Orientālis (talk) 16:56, 3 February 2012 (UTC)

"Spearman's g theory was the dominant theory of intelligence." It still is. Acadēmica Orientālis (talk) 17:10, 3 February 2012 (UTC)

Real-life accomplishments

I tried to find some of the contents of the 'Real-life accomplishments' section in the references provided. Not helped by that no page numbers are given as I have not been able to find any of them in the cites given. Are they actually there does anyone know who has better access than google preview? Dmcq (talk) 14:13, 5 February 2012 (UTC)

I have the book by Kaufman. The page numbers are pages 126 and 132.Acadēmica Orientālis (talk) 14:33, 5 February 2012 (UTC)

Thanks, wanted to check a recent edit and I couldn't. Dmcq (talk) 15:33, 5 February 2012 (UTC)

Removed material

"Heritability" is defined as the proportion of variance in a trait which is attributable to genotype within a defined population in a specific environment. A heritability of 1 indicates that all variation is genetic in origin and a heritability of 0 indicates that none of the variation is genetic. There are a number of points to consider when interpreting heritability. Some examples:

• Heritability measures the proportion of variation in a trait that can be attributed to genes, and not the proportion of a trait caused by genes. Thus, if the environment relevant to a given trait changes in a way that affects all members of the population equally, the mean value of the trait will change without any change in its heritability (because the variation or differences among individuals in the population will stay the same). This has evidently happened for height: the heritability of stature is high, but average heights continue to increase. Thus, even in developed nations, a high heritability of a trait does not necessarily mean that average group differences are due to genes. Some have gone further, and used height as an example in order to argue that "even highly heritable traits can be strongly manipulated by the environment, so heritability has little if anything to do with controllability." However, others argue that IQ is highly stable during life and has been largely resistant to interventions aimed to change it long-term and substantially.
• A common error is to assume that a heritability figure is necessarily unchangeable. The value of heritability can change if the impact of environment (or of genes) in the population is substantially altered. If the environmental variation encountered by different individuals increases, then the heritability figure would decrease. On the other hand, if everyone had the same environment, then heritability would be 100%. The population in developing nations often have more diverse environments than in developed nations. This would mean that heritability figures would be lower in developing nations. Another example is phenylketonuria which previously caused mental retardation for everyone who had this genetic disorder and thus had a heritability of 100%. Today, this can be prevented by following a modified diet which has lowered heritability.
• A high heritability of a trait does not mean that environmental effects such as learning are not involved. Vocabulary size, for example, is very substantially heritable (and highly correlated with general intelligence) although every word in an individual's vocabulary is learned. In a society in which plenty of words are available in everyone's environment, especially for individuals who are motivated to seek them out, the number of words that individuals actually learn depends to a considerable extent on their genetic predispositions and thus heritability is high.
• Since heritability increases during childhood and adolescence, and even increases greatly between 16–20 years of age and adulthood, one should be cautious drawing conclusions regarding the role of genetics and environment from studies where the participants are not followed until they are adults. Furthermore, there may be differences regarding the effects on g and on non-g factors, with g possibly being harder to affect and environmental interventions disproportionately affecting non-g factors.

See no good reason for removing this material so I propose restoring it. Acadēmica Orientālis (talk) 10:38, 6 February 2012 (UTC)

Regression towards the mean is a statistical phenomenon that occurs when an outcome is determined by many independent factors. If an outcome is extreme, then this occurred because most of the independent factors agreed by chance. This is unlikely to occur again so to the next outcome is likely to be less extreme. If IQ is determined by many factors, genetic and/or environmental, then they must mostly agree in the same direction in order to produce an extreme IQ. The child of a person with an extreme IQ is unlikely to have all the factors agree so similarly so the child is on average likely to have a less extreme IQ.

People in professional occupations have on average 25 points higher IQ than unskilled workers. For their children the difference is 21 points. This is in itself not evidence for genetics or environment since the environment for the children likely differs greatly with it on average being more stimulating for the children of professionals.

See no good reason for removing this either. Acadēmica Orientālis (talk) 10:42, 6 February 2012 (UTC)