Gorilla
05-29-2006, 11:22 AM
http://en.wikipedia.org/wiki/Marilyn_vos_Savant
It is generally acknowledged that Marilyn vos Savant has an extremely high IQ. However, there is much confusion over its actual value, with different calculations yielding different numbers: 167+, 186, 218, 228, and 230. These will be examined below, but it should be noted that the measurement of extremely high IQs is an inexact science, subject to problems including small sample sizes (because so few people have IQs at this level), ceiling bumping (because many tests are not designed to measure such high IQs), and a fat tail (because there seem to be more high IQs than a normal distribution would predict), as well as the problems of IQ testing in general.
Marilyn was listed in the 1986 to 1989 editions of the Guinness Book of World Records under "Highest IQ". Subsequent editions do not include this category, and her column now reports that she is listed in the Guinness Hall of Fame. The book mentioned her performance on two intelligence tests: the Stanford-Binet (taken when she was a child) and the Mega Test (taken when she was an adult).
Her Stanford-Binet score is discussed in a 1989 New York magazine article by Julie Baumgold (Baumgold 1989). Marilyn took the Stanford-Binet when she was ten years old; this was the Second Edition of the test, published in 1937. The Stanford-Binet at that time yielded ratio IQs: scores obtained by dividing mental age (as assessed by the test) by chronological age, and multiplying by 100. Marilyn says she first took the test in September 1956, at the age of 10 years and 0 months, and achieved the ceiling mental age of 22 years and 10 months, yielding an IQ of 228. This was the score listed by Guinness, this is the score she gives in interviews, and this is the score shown in the "About the Author" section of her books. Rounding it up produces the value of 230 which sometimes appears.
The figure of 167+ comes from a school record cited by Baumgold indicating that Marilyn took the Stanford-Binet in March 1957, at the age of 10 years and 8 months, and achieved a mental age of "17-10+" (meaning at least 17 years and 10 months). It is unclear how the recorded chronological age was derived; dates in March are 6 or 7 months from her August birthday, not 8. It is also unclear how this record relates to the account given in the previous paragraph. The Stanford-Binet at that time had two forms (Form L and Form M), so one possibility is that Marilyn took the test twice.
The figure of 218 was derived by psychometrician Ronald K. Hoeflin, using a chronological age of 10 years and 6 months, and a mental age of 22 years and 11 months. This figure seems to have no obvious rationale. The ceiling of the Second Edition of the Stanford-Binet was 22 years and 10 months, not 11 months (Terman 1937), and a chronological age of 10 years and 6 months corresponds neither to the age in Marilyn's account nor to the age in the school record cited by Baumgold (although it could fit a March test date).
The second intelligence test mentioned by Guinness is the Mega Test, designed by Hoeflin and taken by Marilyn as an adult in the mid-1980s. The Mega Test yields deviation IQs: scores obtained by multiplying the testee's z-score (the rarity of their raw score on the test) by a constant standard deviation (in this case 16) and then adding 100. Marilyn's raw score was 46 out of 48, corresponding in the latest norming of the test to a z-score of 5.4 and therefore an IQ of 186, a percentile of 99.999997, and a rarity of 1 in 30,000,000 (Hoeflin 1989).
Assertions that Marilyn's IQ dropped from 228 as a child to 186 as an adult are confused: the two numbers represent different types of IQ. For the upper half of the population, ratio IQs seem to follow a log-normal distribution, with a standard deviation of 0.15 for the natural logarithm of the ratio of mental age to chronological age (Scoville). Consequently, Marilyn's Stanford-Binet ratio IQ of 228 corresponds to a deviation IQ of 188, and her Mega Test deviation IQ of 186 corresponds to a ratio IQ of 224.
It is safe to say that Marilyn has one of the highest IQs ever recorded. More extravagant claims—that she is the smartest person in the world, or the smartest woman in the world, or is more or less intelligent than such-and-such a child prodigy, historical genius, or famous intellectual—should be treated cautiously. Marilyn herself values IQ tests as measurements of a variety of mental abilities, but believes that intelligence itself involves so many factors that "attempts to measure it are useless" (vos Savant 2005).
It is generally acknowledged that Marilyn vos Savant has an extremely high IQ. However, there is much confusion over its actual value, with different calculations yielding different numbers: 167+, 186, 218, 228, and 230. These will be examined below, but it should be noted that the measurement of extremely high IQs is an inexact science, subject to problems including small sample sizes (because so few people have IQs at this level), ceiling bumping (because many tests are not designed to measure such high IQs), and a fat tail (because there seem to be more high IQs than a normal distribution would predict), as well as the problems of IQ testing in general.
Marilyn was listed in the 1986 to 1989 editions of the Guinness Book of World Records under "Highest IQ". Subsequent editions do not include this category, and her column now reports that she is listed in the Guinness Hall of Fame. The book mentioned her performance on two intelligence tests: the Stanford-Binet (taken when she was a child) and the Mega Test (taken when she was an adult).
Her Stanford-Binet score is discussed in a 1989 New York magazine article by Julie Baumgold (Baumgold 1989). Marilyn took the Stanford-Binet when she was ten years old; this was the Second Edition of the test, published in 1937. The Stanford-Binet at that time yielded ratio IQs: scores obtained by dividing mental age (as assessed by the test) by chronological age, and multiplying by 100. Marilyn says she first took the test in September 1956, at the age of 10 years and 0 months, and achieved the ceiling mental age of 22 years and 10 months, yielding an IQ of 228. This was the score listed by Guinness, this is the score she gives in interviews, and this is the score shown in the "About the Author" section of her books. Rounding it up produces the value of 230 which sometimes appears.
The figure of 167+ comes from a school record cited by Baumgold indicating that Marilyn took the Stanford-Binet in March 1957, at the age of 10 years and 8 months, and achieved a mental age of "17-10+" (meaning at least 17 years and 10 months). It is unclear how the recorded chronological age was derived; dates in March are 6 or 7 months from her August birthday, not 8. It is also unclear how this record relates to the account given in the previous paragraph. The Stanford-Binet at that time had two forms (Form L and Form M), so one possibility is that Marilyn took the test twice.
The figure of 218 was derived by psychometrician Ronald K. Hoeflin, using a chronological age of 10 years and 6 months, and a mental age of 22 years and 11 months. This figure seems to have no obvious rationale. The ceiling of the Second Edition of the Stanford-Binet was 22 years and 10 months, not 11 months (Terman 1937), and a chronological age of 10 years and 6 months corresponds neither to the age in Marilyn's account nor to the age in the school record cited by Baumgold (although it could fit a March test date).
The second intelligence test mentioned by Guinness is the Mega Test, designed by Hoeflin and taken by Marilyn as an adult in the mid-1980s. The Mega Test yields deviation IQs: scores obtained by multiplying the testee's z-score (the rarity of their raw score on the test) by a constant standard deviation (in this case 16) and then adding 100. Marilyn's raw score was 46 out of 48, corresponding in the latest norming of the test to a z-score of 5.4 and therefore an IQ of 186, a percentile of 99.999997, and a rarity of 1 in 30,000,000 (Hoeflin 1989).
Assertions that Marilyn's IQ dropped from 228 as a child to 186 as an adult are confused: the two numbers represent different types of IQ. For the upper half of the population, ratio IQs seem to follow a log-normal distribution, with a standard deviation of 0.15 for the natural logarithm of the ratio of mental age to chronological age (Scoville). Consequently, Marilyn's Stanford-Binet ratio IQ of 228 corresponds to a deviation IQ of 188, and her Mega Test deviation IQ of 186 corresponds to a ratio IQ of 224.
It is safe to say that Marilyn has one of the highest IQs ever recorded. More extravagant claims—that she is the smartest person in the world, or the smartest woman in the world, or is more or less intelligent than such-and-such a child prodigy, historical genius, or famous intellectual—should be treated cautiously. Marilyn herself values IQ tests as measurements of a variety of mental abilities, but believes that intelligence itself involves so many factors that "attempts to measure it are useless" (vos Savant 2005).