Over the years I’ve made many attempts to estimate the IQs of my readers. Let’s take a look down memory lane:

Gestalt test (Estimated mean self-reported IQ 127 )(U.S. norms)

Based on self-reported performance on a Gestalt test, I estimated my readers have a mean IQ of 123 with an SD of 16 or 19.5 (depending on how Gestalt IQ is calculated). But as a commenter noted, the Gestalt test was normed in Ontario which has a mean IQ of 106 (U.S. norms) and Canada as whole has an SD only 89% as large as America’s (Ontario’s SD is not known).

On a scale where the American mean is set at 100 with an SD of 15, Ontarians are likely 106 with an SD of say,13.4. Using the same U.S. norms, my readers would thus average IQ 127 (SD 14 to 17)

Professionally tested (Mean self-reported IQ 129)(U.S. norms)

Back in May 2017 I befriely ran a comprehensive poll of my readers which included the question What was your most recent score on an official professionally administered individual IQ test taken in the last 10 years? An astonishing 59% claimed they were professionally tested, and their self-reported scores had a mean of 129 with an SD of 21 (U.S. norms). Of course self-reported polls (even anonymous ones) can’t always be taken at face value. There is self-selection in who answers which questions. There are trolls self-reporting IQs as low as 40 or as high as 160+. In addition some of the tests might have been used outdated or non-American norms. Nonetheless, a beautiful bell-like curve emerged:

The SAT (Mean self-reported score equated to IQ 132)(U.S. norms)

Perhaps the single best data-set was from 2015 when I asked readers to select their SAT scores (verbal + math) along with whether they had taken the test before of after the April 1995 re-centering. When converted to IQ equivalents (U.S. norms) readers averaged 132 with an SD of 20. Since 91% of American readers who were old enough to have taken the SAT/ACT had actually done so, those who took the SAT seem representative of American readers as a whole and it’s unlikely that the distribution would change much if the 9% who had not taken it could be forced to do so.

Although the distribution does not form as bell-shaped a curve as the previous graph, there is little evidence of trolling. Till this day no one has ever claimed a perfect score on the pre-recentered SAT. Perhaps this is because the primary troll is “Mug of Pee” and he restrained himself from trolling this one poll out of reverence for the SAT which he considers the best IQ test.

Summary

Gestalt test: Mean IQ 127 SD 14 to 17

Professional IQ test: Mean 129 SD 21

SAT IQ equivalent: Mean 132 SD 20

Average of all three: Mean IQ 129 SD 19

Jonathan Wai method

Professor Jonathan Wai has argued that you can estimate how many top 1% minds are in a group by counting how many elite school alums are in that group. The elite school alums and the top 1% minds will not necessarily be the same people, but the frequency of one arguably predicts the frequency of the other. Assuming the group does not actively select for elite alums, this approach might work.

In 2015, a poll of my readers found that of the 61 respondents who were both American and old enough to be alums, an astonishing 62% claimed to be elite alums implying 62%of my readers have IQs of 135+ (U.S. norms).

I find that hard to believe, but it’s one more piece of evidence pointing to the mean IQ of my readers being somewhere around the +2 SD mark.

The below video is of a Tex Tech student:

According to a comment on Reddit, her SAT score was below 800.

Pumpkinperson.com has not been able to independently confirm that but assuming she took the post 2016 SAT, that would equate to an IQ below 85 (U.S. norms) or below 80 (white norms). It seems the lower the IQ, the less seriously the virus is being taken.

On a scale where American’s average 100 with an SD of 15, Canadian’s have a WAIS-IV full-scale IQ of 104.5 with an SD of 13.4.

Interestingly, American whites score almost as high as Canadians. They have a full-scale IQ of 103.4 with an SD of 14.

If we set the U.S. white mean to be 100 with an SD of 15 (the first IQ tests were normed only on whites), then Canadians have a mean of 101 with an SD of 14.4. So Canadians have virtually identical IQs to American whites. It’s unclear what the IQs of Canadian whites are but I suspect they’re identical to Canadians as a whole.

In my last article I wrote:

Based on the 352 self-reported Gestalt scores, we get the following IQ distribution of Pumpkin Person readers.

We could simply calculate the mean and SD, but given the ceiling bumping (26% got a perfect score) it seems wiser to convert each IQ into a percentile and then a normalized Z score with respect to blog readers.

IQ 74 = 1.13 percentile = -2.27
IQ 90 = 4.39 percentile = -1.73
IQ 98 = 8.225 percentile = -1.4
IQ 106 = 17.6 percentile = -0.93
IQ 114 = 29.68 percentile = -0.53
IQ 121 = 47.15 percentile = -0.07
IQ 129 = 67.04 percentile = +0.47

From here we get the following line of best fit:

y = 122.61+19.546x

The Y intercept = the mean and the slope = standard deviation.

Thus I estimate that on a scale where random whites score 100 with an SD of 15 my readers have a mean IQ of 123 with a standard deviation of 19.5. Such a huge SD is a bit hard to believe but it could be because IQ blogs attract both geniuses and the mentally impaired, thus pulling the curve to both extremes.

It seemed suspicous to me that my readers would have such a huge SD, so I decided to look more closely at the Gestalt test. The Gestalt test was normed using a randomish sample of 14 white Ontarian youngish adults who obtained the following raw scores:

3.5, 3, 2.5, 2.5, 2, 2, 1.5, 1, 1, 1, 1, 1, 0.5, 0.5.

This distribution has a mean of 1.64 and a standard deviation of 0.95 so these numbers were equated to 100 and 15 respectively (the IQ scale). But if the difficulty increase between items is too large or uneven, this might not give a normal distribution, so to be safe, I assigned each raw score a percentile based on where they ranked in the sample of 14 and then equated each percentile to the appropriate IQ on the bell curve:

3.5 = 14/14 = 100 percentile

3 = 13/14 = 92.86 percentile = IQ 122

2.5 = 11.5/14 = 82.14 percentile = IQ 114

2 = 9.5/14 = 67.86 percentile = IQ 107

1.5 = 8/14 = 57.14 percentile = IQ 103

1 = 5/14 = 35.71 percentile = IQ 94

0.5 = 1.5/14 = 10.71 percentile = IQ 81

Armed with these new IQs, I once again found the percentile that each IQ equated to among my readers and the corresponding normalized Z score

IQ 94 = 4.39 percentile = -1.73
IQ 103 = 8.225 percentile = -1.4
IQ 107 = 17.6 percentile = -0.93
IQ 114 = 29.68 percentile = -0.53
IQ 122 = 47.15 percentile = -0.07

This time the line of best fit was:

y = 122.818 + 15.89x

In other words, my readers have a mean IQ of about 123 with an SD of about 16, which sounds more plausible then the absurdly high SD of 19.5 I got before I normalized the Gestalt IQs.

Based on the 352 self-reported Gestalt scores, we get the following IQ distribution of Pumpkin Person readers.

We could simply calculate the mean and SD, but given the ceiling bumping (26% got a perfect score) it seems wiser to convert each IQ into a percentile and then a normalized Z score with respect to blog readers.

IQ 74 = 1.13 percentile = -2.27
IQ 90 = 4.39 percentile = -1.73
IQ 98 = 8.225 percentile = -1.4
IQ 106 = 17.6 percentile = -0.93
IQ 114 = 29.68 percentile = -0.53
IQ 121 = 47.15 percentile = -0.07
IQ 129 = 67.04 percentile = +0.47

From here we get the following line of best fit:

y = 122.61+19.546x

The Y intercept = the mean and the slope = standard deviation.

Thus I estimate that on a scale where random whites score 100 with an SD of 15 my readers have a mean IQ of 123 with a standard deviation of 19.5. Such a huge SD is a bit hard to believe but it could be because IQ blogs attract both geniuses and the mentally impaired, thus pulling the curve to both extremes.

At first glance the idea of Trump being on the autistic spectrum is prima facie absurd. For autistics have impaired social cognition and Trump is one of the most accomplished politicians of our time (despite no political experience, he defeated a platoon of Republicans and the Clinton machine to become President of the United States).

But then Lion of the Blogosphere claimed autism might be in Trump’s family. Well that’s strange I thought. We’d expect a nerdy billionaire like Bill Gates to have an autistic relative, not an alpha-male one like Trump.

In science, it’s such novel and unexpected facts, when fully explored that lead to new understanding. The new understanding might be that autism has nothing to with nerdiness as Lion also had once noted:

I am rejecting an idea I previously had, and which seems common on the internet, that nerdiness is a mild form of Asperger’s Syndrome, or that all nerds have that syndrome. The reason why there is a lot of overlap between nerds and Aspies is because they are both social outcasts, and as social outcasts they both fail to learn the correct behaviors of the popular kids, but the reasons for why they become social outcasts differ. Boys who have a combination of high neuroticism, introversion, small/weak/bad athletic skills, and ugly appearance, are at risk for being excluded from the popular kids groups, and thus become nerds. Unlike Aspies who are often completely antisocial and become interested in weird things like train schedules, nerdy boys are social with other nerds and they have shared nerdy interests like science-fiction television shows and video games. To further confuse things, Aspies often fall into their orbit because the nerdy clique is more welcoming of a weird Aspie than the popular kids clique. And nerds can appear to be anti-social because they are often shy, but it’s not the same cause as Aspie anti-socialness.

Instead of being nerds, Lion noted that autistics lack mimicking instincts.:

Humans are born with a natural instinct to copy the behavior of other humans, and to seek approval of people with higher status in the social group. For people born with autism, this instinct is missing or lacking. For this reason, true autistics fail to learn how to speak because as I’ve previously pointed out, languages are learned by mimicking rather than by reasoning, and learning by mimicking doesn’t happen when a child is born without the instinct to copy his parents.

Partial autistics, in other words those who have Asperger’s Syndrome, do enough mimicking to learn how to talk, but they don’t acquire the fully expressive language abilities of normal children, and they also don’t acquire the correct social behaviors, facial expressions, etc, because those are also learned by mimicking. Lacking the natural copycat instincts of normal children, they usually become social outcasts because they fail to copy the behaviors of the other children and become part of the group. They also tend to pick up odd interests because they lack the copycat instinct of normal children who choose their interests based on what other children like, or what their parents convince them to like.

Aha I thought! This perfectly explains Trump. He lacks the mimicking instincts to copy the virtue signaling of other elites and thus speaks freely about banning Muslims and other immigrants and brags openly about his wealth. His ridiculous hair, trophy wife, orange spray tan, gold plated decorating, and KFC eating shows he lacks the instinct to copy other members of his social class and thus relates better to working class whites than he does other billionares.

As Lion noted, full autistics are so lacking in mimicking that they never learn to speak. Partial autistics (i.e. aspies) do enough mimicking to speak (often at advanced level), but not enough to be political. Perhaps Trump is a partial aspie, so he does enough mimicking to be political, but not enough to fit in with the fine-dining high social class in-which he was born. By failing to mimic the class in-which he was born, the white working class worship him as the one billionaire who is one of them and who will fight for working class interests. By behaving the way most working class whites would if they had billions of dollars (gold plated home, trophy wife) Trump becomes a God-like figure in their eyes.

Another item to add to my regret list: Never experiencing New York city. I did go there when I was eight-years-old (saw a Broadway play Jelly’s Last Jam) but I never experienced it as an adult or even as a teenager. What I remember about NYC was the cold windiness, unbelievable traffic and huge multi-level book stores.

NYC will always have a special place in my heart because it’s the home of the ancient Wechsler Bellevue intelligence scale (which went on to spawn both the WAIS and the WISC series). For David Wechsler, NYC was America and indeed only New Yorkers were included in the test norms of his original scale (though demographically weighted to reflected the country as a whole). You can still see that New York sense of humor in the cartoons the test asks you to arrange and the general knowledge and social judgments it elicits. When I think of American intelligence, I think of New York and there’s no better place to take the WAIS.

But recently New York’s epitaph was written in a brilliant essay called “NYC is dead forever” by hedge-fund manager James Altucher. So powerful was this essay that if NYC was not already dead, this essay killed it. Here, Glenn Beck reads its haunting words :

However Jerry Seinfeld was having none of it, and blasted Atucher for his doom and gloom.

Deviation IQ is defined as one’s Z score (relative to the American or white population) on an intelligence test multiplied by usually 15 and added to 100. Z score = (raw score) – (average raw score)/(standard deviation of raw scores).

Occasionally Z scores will be normalized, so instead of being calculated from the above formula, they’re assigned based on percentiles. So if the top 1% of the population scored X and the normal curve says the top 1% has a Z score of +2.33, X will be assigned a Z score of +2.33 regardless of how many actual standard deviations it is above average.

However not all tests need to be normalized. Some should in theory form a very normal curve on their own. Does anyone know which two of the following tests need to be normalized?:

TEST A

A general knowledge test where the author thinks up general knowledge questions ranging in difficulty from very easy to very hard.

TEST B

A general knowledge test where you have to name the faces of the 100 most important people in history as chosen by Michael Hart’s famous book The 100.

TEST C

A spatial test where the author buys a bunch of jig-saw puzzles from the toy store, and your raw score is the number of puzzles you could put together in under 1 minute each.

TEST D

A spatial test where the author buys one jig-saw puzzle from the toy store and your score is simply how many seconds it takes you to solve it (the lower the better)

TEST E

A spatial test where the author buys one jig-saw puzzle from the toy store and your score is the number of pieces you can fit together in one minute.

TEST F

A memory test where your score is the greatest number of syllables in a sentence you can repeat after one hearing.

If we exclude the 10 known fake votes in the poll of self-reported PATMA scores, 145 people have thus far obtained the following scores:

10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,5,5,5,4,4,3

If you copy and paste this data into this wonderful standard deviation calculator, you get a mean of 7.9 and an SD of 1.4

This distribution forms a gorgeous bell curve though it’s truncated because the test is too easy for about the top tenth of my readership:

On a scale where Nothern American whites average 100 with an standard deviation of 15, my readers are known to average about 126 with an SD around 16. Given this info, there are two ways to convert PATMA scores to IQ.

Method 1: Assign each PATMA score its normalized Z score based on its percentile rank among blog readers and multiply said Z score by 16 and add 126.

Method 2: Calculate the actual Z score of each PATMA score (among blog readers) using the observed mean and SD of PATMA scores (7.9 and 1.4 respectively) and multiply said Z score by 16 and add 126.

PATMA score	frequency among PP readers	percentile rank among PP readers	IQ (method 1)	IQ (method 2)
10	19	93.45	150	150
9	34	75.17	137	139
8	37	50.69	126	127
7	33	26.6	116	116
6	16	9.66	105	104
5	3	3.08	96	93
4	2	1.38	91	81
3	1	0.69	87	70
2	0			59
1	0			47
0	0			36

Method 1 is clearly better. Are any of my readers intelligent enough to understand why?