Americans are scientific geniuses (relatively)

What's up with the poor performance on understanding 25%?

richard posner:
Posner graduated from Yale College (A.B., 1959, summa cum laude), majoring in English, and from Harvard Law School (LL.B, 1962, magna cum laude), where he was first in his class and president of the Harvard Law Review.

now, note this:
In a spirit of self-improvement, I decided to read one of Posner's books. He had one out at the end of last year, as it happens: Catastrophe: Risk and Response. A publisher had sent me a proof copy. (You don't have to buy too many new books in my line of work.) I dived into the Introduction. Page 10:

Suppose there's a one in a thousand chance that the coin when tossed will land on its edge rather than on either of its sides. Suppose further that the coin is tossed only once a year. Then in a thousand years the coin can be expected to be observed on its edge only once.

Hmm. The probability that the coin will be observed on its edge only once in those thousand years is 36.82 percent, which I don't believe rises to the level of an expectation. The probability that it will never land on its edge even once in those thousand trials is almost as great: 36.79 percent. There is an 18.41-percent chance that it will be observed landing on its edge twice, a 6.12-percent chance that it will be thus observed three times, a 1.53-percent chance for four times, and an 0.33-percent chance it will land on its edge five times or more. This is basic probability theory, taking terms from the binomial expansion of (0.999 + 0.001) to the power of one thousand.

Perhaps, as Judge Posner says on page 9 of his introduction: "The human mind does not handle even simple statistical propositions well."

so someone who graduated first in his class at harvard law doesn't grokk the binomial expansion while sagely explaining probability to the masses. so i'll cut the average tard some slack ;-)

"More than half of human genes are identical to those of mice"?

The answer to that should be "disagree". More than half (much more) of human genes have *homologs* in mice (i.e. they share a common evolutionary ancestor), but the number that are absolutely identical in DNA sequence is almost zero.

p.s. what the hey, as long as I'm in correction mode: Razib, an expectation is defined as the number of trials * the probability per trial; 1/1000 per year * 1000 years gives an expected value of one event per year. That is, the mean number of events is 1.0:
which is 0.3679 * 0 events + 0.3682 * 1 event + 0.1841 * 2 events + 0.0162 * 3 events ...
Posner's paragraph is quite correct.

The answer to that should be "disagree". More than half (much more) of human genes have *homologs* in mice (i.e. they share a common evolutionary ancestor), but the number that are absolutely identical in DNA sequence is almost zero.

that's like saying that the big bang isn't an explosion. true, but missing the point of what this test is aiming to measure.

p.s., to be clear: the % of people who will reject the big bang and mouse genetics question because of sophistication is going to be so small that you can ignore them for the purposes of this sort of societal survey.

If Posner is referring to "expectation" in its mathematical sense, then his paragraph is correct.

.3679*0 + .3682*1 + .1841*2 + .612*3 + .153*4 + ... = 1

Whoops, didn't read Sean Eddy's comment.

Becca -

I'm not sure, and I hope Razib will confirm or correct me as appropriate, but I don't think the 0.44 factor relates to how many people understand the meaning of a 1 in 4 probability. I think it's something like a measure of how well their understanding of 1 in 4 predicts their understanding of science overall.

If that's right, I'm not too surprised it's near the bottom of the list. It's a sufficiently simple concept that even people who are quite scientifically illiterate will generally understand it.

I'm not sure, and I hope Razib will confirm or correct me as appropriate, but I don't think the 0.44 factor relates to how many people understand the meaning of a 1 in 4 probability. I think it's something like a measure of how well their understanding of 1 in 4 predicts their understanding of science overall.

*nod* the 'scientific literacy' dimension which can account for most of the variation in response on these questions....

(though from what i recall the raw scores are pretty shocking, e.g., a substantial minority of americans are geocentrists)

that's like saying that the big bang isn't an explosion. true, but missing the point of what this test is aiming to measure.

Good point, and you could do that with a lot of these. For example, "The earliest humans lived at the same time as dinosaurs," is a true statement under phylogenetic nomenclature, wherein Aves is a clade within Dinosauria.

i think the electrons & atoms question might be debatable too because of the weirdness of something like "size" at that scale (not debatable if you say mass of course).

Oh! That makes much more sense. I guess I'm relieved... though I'm still a little unclear why it should correlate so poorly. I understand the politically charged ones- scientifically literate folks basically choose to believe something other than scientific sources would support.
@ Mike Keesey-
While I agree with Razib that these likely technicalities probably don't matter on a population level, I take a perverse sort of joy in finding ways the "correct" answers are incorrect (more broadly than in this sort of quiz). I'm glad I'm not totally alone.

I take a perverse sort of joy in finding ways the "correct" answers are incorrect (more broadly than in this sort of quiz).

i assume readers of this weblog are much more likely to find precise errors or conflations in these sorts of formulations. science has to push the margins of generality; i.e., newtonian mechanics is "wrong" too on the most general level.

That's pretty weird re: "expectation." It seems clear that's what Posner meant. Maybe "mean of binomial = n*p" is like the multiplication tables to Derbyshire, who trained in mathematics; just something you forget after cramming other stuff in its place.

Becca,

I look at it this way. Suppose the only thing I know about Person A is that they can correctly define 'stem cell.' The only thing I know about Person B is that they have a correct understanding of the meaning of a 1 in 4 probability.

Knowing only that, if I had to predict which could correctly define 'molecule' or 'DNA', it seems obvious to me that Person A is more likely than Person B.

What do "open-ended" and "closed-ended" even mean?

Sigh, i'm moving to Sweden

What do "open-ended" and "closed-ended" even mean?

close-ended -> agree/disagree, true/false. open-ended are "in your own words" explanations.

Agnostic: see http://en.wikipedia.org/wiki/Expected_value for the mathematical details. Put simply, expected value, mean, and integral are synonyms. This means that if you repeat the one-hundread-years-experiment several times (yes, that's long) and average the results, the average will be one.

Besides, the value « one » is also the mode of this particular statistical experiment. So the sentence is true in that sense, too.

-- a passing mathematician

The difference between "homologous" genes and "identical" genes isn't technical nitpicking; it's a serious enough misunderstanding on the part of the folks who wrote the question that they'd better not be making any snarky comments about lay people's (mis-)understanding of modern genetics. Not understanding the difference means not understanding molecular evolution.

It's essentially the same as asking, true or false, more than half of Spanish words are identical to words in Italian? False, even though most Spanish words clearly share recent common ancestry with Italian words.

sean, again, you're missing the point. the purpose of this survey is to get good indicators of general science knowledge. introducing words like "homologous" into the survey would basically mean it became unintelligible. the press routinely reports these "facts" with words like "identical." the judge of 'scientific literate' is someone who can read the science section of the new york times, not someone who could navigate molecular biology of the gene.

the big bang isn't an explosion either. but most people don't get space-time curvature, so how the hell are they going to comprehend expansion of space?

and it isn't the same thing as language. people share many referents toward understanding language.

btw, you know there are also disputes on definitions of words like "genes." it really never ends.

and let me offer an argument to sean why technical precision isn't really the best for this survey. someone who agrees that the big bang is an 'explosion' is wrong by how we understand explosions in a precise sense. but, if they think that the big bang was an explosion they obviously have something of a general idea of the big bang translated into their common everyday referents. similarly, the mouse gene question is querying the general idea that organisms share common genetic substrate, and that the similarity of that substrate varies in proportion to phylogenetic distance. many of my creationist friends rejected these sorts of assertions in shock when i initially made them in high school because it didn't make sense in their paradigm (for obviously reasons). this is fundamentally social science based research, so the nature of these natural science questions is instrumental, they aren't the ends.

Most Europeans don't know what an inch is. Did they really ask question 3 in Europe? I'm missing something... that would be kind of culturally biased, wouldn't it?

Having knowledge of a scientific term or idea is not the same as being scientificaly literate. It is knowledge & utilisation of the scientific method not scientific discoverys that makes a person scientifically literate. Are the Iranians scientifically advanced for knowing about uranium?

It's curious that Nuclear power plants destroy the ozone layer gets a loading of 0.74, whereas Nuclear power plants destroy the ozone layer gets a loading of only 0.68.

sean, again, you're missing the point. the purpose of this survey is to get good indicators of general science knowledge. introducing words like "homologous" into the survey would basically mean it became unintelligible. the press routinely reports these "facts" with words like "identical." the judge of 'scientific literate' is someone who can read the science section of the new york times, not someone who could navigate molecular biology of the gene. the judge of 'scientific literate' is someone who can read the science section of the new york times, not someone who could navigate molecular biology of the gene.

They could easily use "counterpart" like this NY Times story on the mouse genome. I didn't find any claims of "identical" human & mouse genes there.

similarly, the mouse gene question is querying the general idea that organisms share common genetic substrate, and that the similarity of that substrate varies in proportion to phylogenetic distance.

How is it testing the latter?

Heh... This reminds me of one of the least scientifically literate guys I've ever known. He thought that "stem cells" were called that because they were taken from the brain stems of babies. This was his reason for opposing stem cell research. He also refused to believe me when I told him he was wrong and explained to him what stem cells really are. He just insisted that he was right and that I was the one who was mistaken. I finally had to make him read the Wikipedia page on stem cells before he would believe me.

Knowing that guy, it would not surprise me at all if he's gone back to his incorrect definition of the term. Ignorance like that is scary.

How is it testing the latter?

there's usually a chimp question in there too from what i recall.

Yep, but you can't get much info from comparing the chimp and mouse questions in this survey. The cutoff points are too much alike ("more than half"/"less than half").

Apparently I can get many of these questions wrong and still be "scientifically literate?" The "scientifically literate" author doesn't provide a clear and straightforward explanation of what was done, but can only refer you to a few dozen other papers for your reading pleasure. Maybe in one of these other papers you might be able to figure out whether the survey itself was used to establish "scientific literacy" (if so then what were the actual questions and what was a "passing" grade?) or if it is an indicator (per each loading factor) of a scientific competence established by some other means (such as possessing a degree or certification)? I have no idea whether the questions of table 1 constitute the entire survey or some "interesting" subset or a set of follow-up questions or why so many politically and religiously charged questions were included.
I think the basic requirement for such a paper should include an explanation what was done and a summary of the raw data.

Most of these questions are no more than memorization of (albeit poorly worded) "facts", like memorizing history dates.

It probably does not measure scientific literacy understanding-wise (ie. the scientific method, experimental design) , only exposure to news headlines and catchy scientific statements one hears in the popular press.

Am kommunistpartiet skrev en bok "We will never surrender" (hette = men historisk analys av gamla österrike ungern, såldes i få exemplar )

tillägg

Vi överlever inte utan nyliberalism (am projektet= upplysning, kraft, eget ansvar etc).
Sex, våld och arbete.
Sen är det den judiska frågan - många judar inom arbetarrörelsen (säljer in intellektuella lösningar som kommunistpacket idag "massa meningslösa ord utan substans").
Samma mentalitet idag = kanske att naturen går mot utrotning. Kontroll över en del media - och liknande organisationer.
Djävulens folk (enkelt å slå ut).

Räkna 1 miljon gånger
Vad som gett nybyggaranda etc

Spindel som kan döda dom

Det var ju en judisk flicka som nu heter Rebecca Lord
Hon var inte kåt men väl listig å manipulera andra (Rebecca mata hari).
Hon blev påsatt och nedslagen men en dag begick hon självmord.
Sen kom en klumpedunsröv från israels högre uni , Haifa, och skulle lära sig empirism men då slog schizopartikeln till och det kliade till i röven och han drog ut en hemorojd
den visades sig innehålla en mask som sedemera förintade en stor del av den mänskliga världen

det stod sen i historieböckerna
astroillium israelis

partikel =
nervsystemet har ju nya rön om databasförmåga (aktiverar då schizofreni i oss)
Tex sända nanorobotar och religiösa kulturella kontexten och spela på (socialpsykologi) och då äventyra mänskligheten
hot mot mänskligheten

när man sitter i en isoleringscell (med 1000 mdr som egentligen skulle gått till väst-am projektet) och attackhund som våldtar och våldsmusik

kinesisk han-kines som sa
nu erövrar vi manchuriet
så kom hästarmén som var genklonad
och man sålde sin kossa
det visade sig sen att han var lite underskattad
så man impoterade en ko

(vv sprid vidare)

Det var ju en judisk flicka som nu heter Rebecca Lord
Hon var inte kåt men väl listig å manipulera andra (Rebecca mata hari).
Hon blev påsatt och nedslagen men en dag begick hon självmord.
Sen kom en klumpedunsröv från israels högre uni , Haifa, och skulle lära sig empirism men då slog schizopartikeln till och det kliade till i röven och han drog ut en hemorojd
den visades sig innehålla en mask som sedemera förintade en stor del av den mänskliga världen

det stod sen i historieböckerna
astroillium israelis

ang kulturarvet= vi borde förvalta de kulturella rön om tex kristna empiri (ulf danielssons bok tex om himlakroppars och rörelse i tiden = läsa av), eng matroserna och röda blodslinjen och aliens spår (från konstantin till maya och hit och dit)
Tolkiens briljans (prekognition ww1 om vad han skulle skriva)

ang psy krigsföring (gamla tillbaka)
ingenting å skringra kina (sätta in dödsstöten)
kusterna = stänga in
små nanokapslar av det kobröstmjölk som dom vill sälja till oss

Am kommunistpartiet skrev en bok "We will never surrender" (hette = men historisk analys av gamla österrike ungern, såldes i få exemplar )

tillägg

Vi överlever inte utan nyliberalism (am projektet= upplysning, kraft, eget ansvar etc).
Sex, våld och arbete.
Sen är det den judiska frågan - många judar inom arbetarrörelsen (säljer in intellektuella lösningar som kommunistpacket idag "massa meningslösa ord utan substans").
Samma mentalitet idag = kanske att naturen går mot utrotning. Kontroll över en del media - och liknande organisationer.
Djävulens folk (enkelt å slå ut).

Räkna 1 miljon gånger
Vad som gett nybyggaranda etc

Spindel som kan döda dom

Det var ju en judisk flicka som nu heter Rebecca Lord
Hon var inte kåt men väl listig å manipulera andra (Rebecca mata hari).
Hon blev påsatt och nedslagen men en dag begick hon självmord.
Sen kom en klumpedunsröv från israels högre uni , Haifa, och skulle lära sig empirism men då slog schizopartikeln till och det kliade till i röven och han drog ut en hemorojd
den visades sig innehålla en mask som sedemera förintade en stor del av den mänskliga världen

det stod sen i historieböckerna
astroillium israelis

partikel =
nervsystemet har ju nya rön om databasförmåga (aktiverar då schizofreni i oss)
Tex sända nanorobotar och religiösa kulturella kontexten och spela på (socialpsykologi) och då äventyra mänskligheten
hot mot mänskligheten

när man sitter i en isoleringscell (med 1000 mdr som egentligen skulle gått till väst-am projektet) och attackhund som våldtar och våldsmusik

kinesisk han-kines som sa
nu erövrar vi manchuriet
så kom hästarmén som var genklonad
och man sålde sin kossa
det visade sig sen att han var lite underskattad
så man impoterade en ko

(vv sprid vidare)

mkt skitsex o sånt

en golem full av skit som sen formerar sig som en dildo (en israelisk uppfinning från 1934, Hitler som köpte den och sen gav han himalaysiska presidenten en bil där det fanns 20 sådana i bakluckan)
han hade ju också en dvärgchaufför och såg en svart spökfigur springa omkring enligt påven

http://juridiskaresonemang.blogspot.com/

Sweden is not how you spell Sweeden KY