btw, about those victorians…

in my post about the woodley, et al., iq paper the other day, i mentioned that one of the two studies on victorian iqs included in their research had been conducted on university of chicago students. i got to thinking afterwards that, apart from the fact that most of those students had probably been white (probably wasps, in fact), this could hardly have been a representative sample of victorian americans since the vast majority of people didn’t attend college in those days — late nineteenth century college students would’ve mostly been, you know, actual smart people.

so, then i was planning on (reluctantly) splashing out $11.95 on this article to find out the details on what sorts of people were included in galton’s study of victorian english iqs, the second set of victorian iq results used by woodley, et al. (reluctantly because, h*ck! — $11.95 is three or four tall decaf mochas!).

but scott alexander over @slate star codex is way ahead of me! scott writes:

“Galton’s Data A Century Later, published in 1985, tells us a little about how he gained his ground-breaking reaction time statistics. He set up a laboratory in the Science Galleries of the South Kensington Museum. There he charged visitors to the museum three pence ($25 in modern currency after adjusting for inflation) to be measured by his instruments, a process he advertised as ‘for the use of those who desire to be accurately measured in many ways, either to obtain timely warning of remediable faults in development, or to learn their powers.'”

ehhhh. charged a fee? uh-oh.

scott acknowledges that his $25 estimate mightn’t have been the most accurate, so i decided to use the calculator over at measuring worth to see how much three pence from 1889 would be worth today (i used their average earnings index), and i got £5.25 or $7.94 (for 2010). that’s not quite $25, but still that’s two or three tall mochas! i’m not sure that very many lower class victorians would’ve been willing or able to part with that amount of money just to take galton’s funny little test.

additionally, there may also have been an admission fee for the south kensington museum (now the victoria and albert museum) that we might have to add to that $7.94, but i’m not sure because i couldn’t find out anything about museum fees in victorian era london. (i know, for instance, that museums in places like new york generally had fees in the late 1800s, but also often had one free day a week to enable the lower classes to visit and learn. bunch o’ idealists, the victorians! (~_^) )

so again, i think we’ve got an apples and oranges problem here.

go read the whole of scott’s post, btw, because it’s excellent!

previously: we’re dumber than the victorians

(note: comments do not require an email. the south kensington museum.)

Advertisements

24 Comments

  1. If he collected on a “free-to-the-public day,” he may have chosen to do so knowing that it would potentially provide a wider data sample. Did it? Who knows.

    /Not backing the current study, btw. I haven’t read it in detail yet.

    Reply

  2. @hbd chick – I’m afraid you are, inadvertently, doing here with the Woodley at al paper exactly what the IQ police always do with work they don’t want to believe.

    That is to say, you are undertaking an absurd and unique level of micro-methodological analysis of the disbelieved IQ research in a way which is *never* done with believed-in IQ research – in this case sampling methods.

    Have you ever asked about any other IQ paper how much volunteers were paid? No. So this micro-methodological quibbling is not about doing science, it is not about trying to discover what is going-on in the real world – it is in fact looking for excuses to reject a paper.

    Don’t worry! – nobody can *make* you believe anything; every single piece of scientific research ever done on the planet has been and is open to other interpretations and is imperfect!

    But you must not engage in distorted analysis to rationalize the fact that you don’t want to believe it.

    *

    Is the apparent lengthening of reaction time over the past century plus interesting? You bet it is!

    Is the Woodley at al paper up to the prevailing standards of research competence? You bet it is.

    Is Woodley an honest researcher driven by a desire to discover the truth? You bet he is (something which cannot be said about many modern scientists).

    So – should the Woodley et al paper be ridiculed and rubbished and rejected without further consideration, on the basis that there might, perhaps, be other explanations for the data? Of course not!

    *

    I was an lecturer in epidemiology for three years, and have published extensively on the matter of sampling – and I notice such matters.

    I notice that IQ researchers almost never notice or discuss sampling issues – for example, http://iqpersonalitygenius.blogspot.co.uk/2012/08/how-should-we-measure-general.html

    So, why the sudden interest in micro-sampling issues, from people who have shown no interest (and show little knowledge or competence) in this issue yet suddeny being applied to Woodley at al’s paper?

    What this looks like is an attempt to discredit the Woodley et al paper, and a rather foolish attempt to use sampling issues as the handiest tool to achieve this aim.

    Because if this micro-level sampling critique was applied to the whole IQ literature (rather than just to Woodley at al) there would be almost nothing left!

    Ergo, it is a fake critique – because people are arguing from a false premise based on ignorance of sampling issues in the mass of IQ research; as if ignorance of sampling issues in other papers, due to not knowing and asking the question, was equivalent to better sampling!

    *

    The Silverman paper states clearly, and I said this in my original blog post and Woodley at al says it, that Galton’s reaction time data is confirmed by several other contemporary measures from independent workers.

    “Silverman notes that in reviews of reaction time studies in 1911 (but not including Galton’s work), it is clear that Galton’s results were typical of the era – the range being from 151-200 milliseconds – median of 192 milliseconds.”

    http://charltonteaching.blogspot.co.uk/2012/02/convincing-objective-and-direct.html

    Also, the modern data can be broken out to reveal a matched sample of museum visitors, which shows the same magnitude of increase in RT as the overall meta analysis.

    *

    So, let’s assume that the sampling critique of Woodley’s paper is wrong, and the measured increase in RT cannot *plausibly* be explained by sampling differences (unless we are prpeared to reject almost every paper on IQ research ever published because of failure to deal with micro-level sampling issues) – and therefore the measured increase in RT is *probably* real: what then?

    *

    The next step is to ask whether the *implications* of a slowing of RT (and decline of g) can be be tested, and perhaps refuted, using other evidence of comparable precision and objectivity to reaction times?

    The way we do this is that we assume (contingently, as a way of testing this assumption) that IF reaction times have slowed significantly (and/ or g has reduced significantly), THEN we would expect to observe X, Y or Z – and then we try to ascertain whether we do IN FACT observe X, Y and Z.

    Or we could say that IF we can observe A, B or C – then decline in g CANNOT – have happened – then we go out and look-for A, B or C.

    That process is TESTING THE PREDICTIONS derived from Woodley et al: and that is how real science works!

    And this is what needs to be done next.

    *

    Reply

  3. Mr. Charlton with all due respect that’s the weakest and most rambling comment I’ve seen in a blog in quite some time. You aren’t convincing anyone writing like that.

    Reply

    1. @spandrell – I don’t really care one way or another about ‘convincing’ some random pseudonymous commenter of unknown competence and honesty – I am talking to hbd chick. If you don’t understand what is being said – then clearly this is not for you.

      Reply

  4. I don’t think the criticism of the race issue is valid. Yes the population includes a lot of non-whites now, and that will affect the results of your test. That’s not sampling bias, that’s the reality..

    Reply

  5. @bruce – first of all, the personal stuff:

    do not — i repeat do NOT — leave any more comments on this blog like the snarky ones above in which you insult myself or other commenters with an accusatory tone. make sure your next comments are measured and civil, or i will ban you from my blog altogether (something which i’d rather not do since i have, in actuality, a lot of respect for your blogging and your past work at Medical Hypotheses — notice that i have you in my blogroll).

    also, before you fling completely unfounded accusations at other bloggers, you would do well to actually READ what they’ve written.

    for instance, you say about me:

    “But you must not engage in distorted analysis to rationalize the fact that you don’t want to believe it.”

    where on EARTH did you get that idea from? did you bother to READ both of my posts on the woodley, et al., paper — the first of which i linked to THREE times in this post? obviously not, otherwise you would have noticed that i closed with the following:

    “this is a really neat study, and maybe woodley, et al., are on to something, but i’d like some reassurance that they’re looking at the same sorts of populations.”

    does that SOUND like i don’t want to believe that iqs have gone down in the west since the victorian period? no.

    i’m not going to bother addressing all of the other slanderish notions you seem to have about what i think (there are too many), because they are all as baseless as this one. but, again, i suggest you have a read through someone’s blog before you make ASSUMPTIONS about what they think and what their motivations are. or you could just ask me what i think — i’d be happy to tell you.
    _____

    @bruce – now, the scientific stuff.

    @bruce – “Have you ever asked about any other IQ paper how much volunteers were paid?”

    no, but i’m not a psychometrician, nor is iq my main interest in human biodiversity. however, again if you had bothered to read my blog, yes i HAVE questioned the particulars wrt to sampling when it comes to theories about iq — see here for example: more from ron unz on iq.

    @bruce – “Is Woodley an honest researcher driven by a desire to discover the truth?”

    i never said, or even suggested, that he’s not.

    @bruce – “I notice that IQ researchers almost never notice or discuss sampling issues….”

    well, they ought to!! this is basic, basic — i’m talking really fundamental — science 101: to make sure you’ve done your sampling right. and if you want to find out if, and by how much (if the answer is yes), iqs have dropped in ANY population over time, you’d better make sure you’re looking at the same population at both ends of your study!

    @bruce – “The Silverman paper states clearly, and I said this in my original blog post and Woodley at al says it, that Galton’s reaction time data is confirmed by several other contemporary measures from independent workers.

    ‘Silverman notes that in reviews of reaction time studies in 1911 (but not including Galton’s work), it is clear that Galton’s results were typical of the era – the range being from 151-200 milliseconds – median of 192 milliseconds.'”

    unfortunately, i don’t have access to the silverman paper, but i understand that silverman is refrencing ladd and woodworth’s Elements of Physiological Psychology (1911), which is available online here.

    you and woodley, et al., say that silverman says that ladd & woodworth’s review of other victorian studies done on rts found rts in the range of 151-200 ms. i’ve looked through ladd & woodworth and found a table on page 476 which appears to be what silverman was referring to (again, i can’t be sure because i don’t have the silverman paper). that table lists eight studies with a range of rts to optical stimuli of 151 (0.1506) to 225, not 200. not sure why silverman left that 225 study out (the hankel study), if that is in fact what he did — perhaps he explains his reasons in his paper.

    in any case, while the existence of these other studies is great and their results fascinating, i would still want to know who was included in those studies! because, again, sampling is actually very important and shouldn’t be overlooked or ignored. do you know who was included in those studies? does woodley, et al.? i would hope so!

    as scott alexander pointed out about the galton data:

    “The 1985 commentary on Galton’s data tells us:

    “‘As would be expected of a group of paying testees being measured in a museum, a sizable portion of Galton’s sample consisted of professionals, semiprofessionals, and students. However, as may be discerned in Tables 10 and 11, all socioeconomic strata were represented.’

    “Tables 10 and 11 turn out to be a gold mine – I worried the records of exactly who took the tests would be lost, but as you might expect of someone who basically invented statistics single-handedly and then beat Darwin in a debate about evolution as an encore, Galton was very good at keeping careful data.

    “This site tells me that about 3% of Victorians were ‘professionals’ of one sort or another. But about 16% of Galton’s non-student visitors identified as that group. These students themselves (Galton calls them ‘students and scholars’, I don’t know what the distinction is) made up 44% of the sample – because the data was limited to those 16+, I believe these were mostly college students – aka once again the top few percent of society. Unskilled laborers, who made up 75% of Victorian society, made up less than four percent of Galton’s sample!”

    so, my questioning of the costs for taking galton’s test, and even entry to the south kensington museum, is not some crazy exercise in nitpicking. these could have — and appear to have actually — skewed galton’s sample.

    @bruce – “That process is TESTING THE PREDICTIONS derived from Woodley et al: and that is how real science works!”

    science ALSO works by the open discussion of findings and methodology, NOT by the shouting down of anyone who might have questions regarding either of those.

    and science works, too, by scientists attempting to replicate each other’s research in order to confirm whether or not their findings hold water, a process which i hope happens in this case because, although i know you find this hard to believe, i do in fact want to know if western iqs have sunk since the victorian period.
    _____

    take care in how you word your next comment.

    Reply

  6. @daybreaker – “I don’t think the criticism of the race issue is valid. Yes the population includes a lot of non-whites now, and that will affect the results of your test. That’s not sampling bias, that’s the reality.”

    the point is, woodley, et al., are arguing that iqs have gone down in the western world since the victorian age due to dysgenics. maybe. but they haven’t taken into account either 1) the sampling biases in the victorian studies (possibly making the average victorian iq appear to have been higher than it was), or 2) the presence of non-whites in the modern day samples (possibly making the average modern iq appear to be lower than it is).

    Reply

  7. @hbd chick – You can, of course, conduct your blog any way you want – I tend to filter out all comments that I don’t want to print – maybe replying by e-mail. But here I was treating you as a serious scientist who had made a serious error in the way you are evaluating *potentially* paradigm-changing research. I was hoping you would recognize this error, once it was pointed-out, and that you would repent. Seemingly this is not going to happen. Let’s leave it at that.

    Reply

  8. @bruce – “Let’s leave it at that.”

    yes, i guess we’d better had, since you don’t seem to think that sampling methodologies in science are important.

    Reply

  9. brucecharlton
    “But you must not engage in distorted analysis to rationalize the fact that you don’t want to believe it.”

    Personally i think it’s highly likely average IQ among the indigenous population in England peaked before the industrial revolution and has been gradually dropping since due to the side-effects of industrialization and separate from any immigration effects. I think the hesitation among HBD people is mostly over the scale of the drop.

    Reply

  10. @grey – “I think the hesitation among HBD people is mostly over the scale of the drop.”

    exactly. and i think that, perhaps, some of those 15 points might possibly be accounted for by woodley, et al., overestimating the victorian iq and underestimating today’s iq thanks to sampling biases.

    dr. james thompson over @psychological comments said that woodley & co. will probably address these issues in a post on his blog. i look forward to reading it! maybe they have good reasons for not being concerned about the sampling.

    (i’m sure i’ll much prefer reading their response than being browbeaten by bruce charlton … right here on my own blog! =/ )

    Reply

  11. Bruce Charlton badly, badly wants to believe that the modern day Briton has gotten roughly one stdv dumber since the Victorian era, and that this reflects a purely genotypic decline. He so badly wants to believe in it that he makes a fool out of himself through specious reasoning and insults random people who point out flaws in his evidence.

    As I pointed out on Cochran’s blog, in order for a 1 stdv decrease in IQ over the past 120 or so years = ONLY the dumbest 60-70% would have to breed with each generation….. Just what the HELL? It has also been pointed out that reaction times correlate poorly with IQ, and so that a decline in one may indeed have nothing to do with the other – a timeless, unbiased measure of intelligence, they are anything but.

    And no, we are not going to dismiss a biometric model that has worked since Galton’s day merely to satisfy Bruce Charlton’s personal prejudices. A genotypic decrease of 1 stdv in *any* phenotype over the span of 120 years, even one with a narrow sense heritability as high as height, would require a selection differential greater than the impact of the Black Plague on the population of medieval Europe. You tell me how stupid and crazy that sounds. It’s not even remotely plausible.

    Reply

  12. “‘Silverman notes that in reviews of reaction time studies in 1911 (but not including Galton’s work), it is clear that Galton’s results were typical of the era – the range being from 151-200 milliseconds – median of 192 milliseconds.'”

    if anyone’s interested in actually checking what sorts of populations were included in these studies reviewed in 1911 (it’s not my research, so it’s not my job!):

    the 1911 review is ladd and woodworth’s Elements of Physiological Psychology, which is available online here. see the table on pg. 476.

    ladd and woodworth got that table from kries and auerbach (1877) — pg. 359 here. note that the publication date is 1877, so all of these studies are at least seven years older than the start of galton’s study.

    the first one on the list is hirsh which, i figured out, is Chronoskopische Versuche über die Geschwindigkeit der verschiedenen Sinneseindrücke und der Nerven-Leitung, a talk given in 1861, so a full 23 years before galton started his work. i couldn’t see anything about who the subjects were in hirsh’s studies, but my german is very rusty (i.e. practically nonexistent!).

    i think a lot of this nineteenth century research might be found at this max planck virtual library site. it shouldn’t be impossible to double-check these early samples.

    Reply

  13. “(i’m sure i’ll much prefer reading their response than being browbeaten by bruce charlton … right here on my own blog! =/ )”

    aww, quite :)

    to take your mind off it

    Reply

  14. (Still not getting this…) But If non-whites (and mixed-race individuals) are occupying the West in large numbers now, which they are, particularly in the big cities (which is where I would expect the sampling to be done), that doesn’t appear to affect the average IQ of individuals occupying the West, it actually affects it. (And the culture too, and not necessarily in ways that are good for active IQ.)

    I mean, if the Shire gets occupied by orcs, then any psychometric testing you do before and after should tell you the average inhabitant if the shire has become much more orc-like, because he has.

    In fact if you have a normal bias to testing where it’s convenient and where academic people like to be, you’ll ignore all the hobbits toiling in their fields and do your tests in the town square, where the orcs are strutting around cracking their whips and eating up the hobbits’ food. Is that it, are you talking about an urban / rural bias in testing?

    Because otherwise I’m not getting this.

    Reply

  15. @grey – “to take your mind off it”

    awwww! so cuuuute! (^_^) (i’m such a girl.) thanks!

    whenever i’m feelin’ kinda blue, i have a look at this guy [<< WARNING!: cute overload] — despite his poor broken foot, he's so sweet, can't help but (^_^).

    Reply

  16. There are definitely reasons to think a genotypic IQ decline over the past 150 years is very plausible – dysgenics and low IQ immigration, the decline of innovation, etc. But a 1 standard deviation/15 point drop does seem extremely high; the equivalent of replacing all Europeans with westernised Africans, and much more than could be achieved on any likely dysgenic theory. I’d expect more like a 5-7 point drop.
    .

    Reply

  17. @daybreaker – “(Still not getting this…) But If non-whites (and mixed-race individuals) are occupying the West in large numbers now, which they are, particularly in the big cities (which is where I would expect the sampling to be done), that doesn’t appear to affect the average IQ of individuals occupying the West, it actually affects it.”

    i would think so. to some degree or another (but that would depend upon how many immigrants you had in your society — OR possibly how many immigrants you had in your sample that you’re looking at).

    what woodley et al. are arguing is that the 14.1 point drop in western iqs that they think that they’ve found is due primarily to dysgenics — i.e. that lower iq westerners have been reproducing more, and their offspring surviving in greater numbers, since the victorian period. maybe. but i think they’ve got some sampling issues, and that that might be obscuring a lower, but more realistic (like simon suggests above) drop of ca. 5 points. maybe. dunno. pure speculation.

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s