When I revised my formula for converting post-April 1995 SAT scores to IQ equivalents, it meant that I needed to also revise a chart I made converting SATs to IQ. Below is the revised chart:
sat score (verbal and math combined) | pre-april 1995 iq equivalent (u.s. norms) | pre-april 1995 iq equivalent (u.s. white norms) | post-april 1995 iq equivalent (u.s. norms) | post-april 1995 iq equivalent (u.s white norms) |
425 | 62 | 59 | 58 | 53 (Trainable Moderate Mental Retardation) |
475 | 69 | 66 (Educable Mild Mental Retardation) | 62 | 58 |
525 | 74 | 71 | 66 | 62 |
575 | 80 | 77 (Borderline Mental Retardation) | 70 | 66(Educable Mild Mental Retardation) |
625 | 84 | 81 | 74 | 70 |
675 | 89 | 87 (Dull) | 79 | 75 |
725 | 93 | 91 | 83 | 79(Borderline Mental Retardation) |
775 | 96 | 94 | 87 | 83 |
825 | 106 | 104 (Average) | 91 | 88 (Dull) |
875 | 108 | 106 | 95 | 92 |
925 | 111 | 109 | 99 | 96 |
975 | 115 | 113 (Bright) | 103 | 100 (Average) |
1025 | 118 | 117 | 107 | 104 |
1075 | 120 | 119 | 111 | 108 |
1125 | 123 | 122 (Very Bright) | 115 | 112 (Bright) |
1175 | 127 | 126 | 119 | 117 |
1225 | 131 | 130 (Extremely Bright) | 123 | 121 (Very Bright) |
1275 | 133 | 132 | 127 | 125 |
1325 | 137 | 136 | 132 | 130 (Extremely Bright) |
1375 | 140 | 139 | 135 | 133 |
1425 | 145 | 144 (Brilliant) | 139 | 137 |
1475 | 149 | 149 | 143 | 141 (Brilliant) |
1525 | 155 | 155 (Very Brilliant) | 147 | 146 |
1575 | 165 | 165 (Extremely Brilliant) | 151 | 150 (Very Brilliant) |
perfect 1600 | 171 | 171 (Mega Level) | 153 | 152 |
The above table contains columns which show IQ equivalent expressed on a scale where the mean and standard deviation of the overall U.S. population is set at 100 and 15 respectively (U.S. norms) and on a scale where the overall U.S. white population is set at 100 and 15 respectively (U.S. white norms).
To determine the U.S. norms IQ equivalent for post-April 1995 SAT scores, I used this REVISED formula which I created:
IQ equivalent = 23.835 + 0.081(new SAT score)
To determine the U.S. norms IQ equivalent for pre-April 1995 SAT scores above 789, I was able to get approximate numbers using WAIS IQ equivalents in a chart in section 8.3.3 of the Prometheus Society Membership Committee report.
To determine U.S. norms IQ equivalents for pre-April 1995 SAT scores below 790, I converted scores on the original scale to scores on the recentered scale, and then converted to IQ using the REVISED formula I created, mentioned above.
To convert U.S. norms IQ equivalents to U.S. white norms IQ equivalents, I assumed that before April 1995, on a scale where all Americans averaged 100 (SD = 15), white Americans averaged 102 (SD = 14.5). After, April 1995, I assumed (citing data from Charles Murray’s book Coming Apart) that on a scale where all Americans average 100 (SD = 15), white Americans average 103 (SD = 14.5).
Using this data, I converted U.S. norms IQ equivalents into Z scores with respect to the U.S. white distribution of their time, and then multiplied said Z scores by 15 and added 100, to get the U.S. white norms IQ equivalents.
Because IQ is scored on a bell curve, and whites generally score higher than most minorities on standardized tests, the more non-white America becomes, the further above 100 the white IQ goes. To avoid this statistical artifact, it’s best to use U.S. white norms which keep the mean and standard deviation of white Americans at 100 and 15 respectively, while the mean and SD of America as a whole is allowed to vary. See here for more details.
Lastly, the above chart is not intended to predict your score on an official IQ test based on your SAT score. Instead, it is simply your SAT score expressed on the same scale as IQ. The correlation between SAT scores and official IQ tests is far from perfect, so people with high SAT scores often regress to the mean on official IQ tests, and vice versa. To see just how acute the effects of regression can be, see the IQs of Harvard students.
your formula is shit peepee.
the prometheus society membership ALL have LOW IQs.
so weird how this has only three comments and the other one has 108
SAT scores never end in a five, only in 0.
Hey pumpkinperson, great post, but why not simplify and use Maine as a reference for converting sat into iq, they are still very homogeneous about 95% white according to Wikipedia, latest figure 2016: https://pt.m.wikipedia.org/wiki/Maine and its mostly Anglo-whites which according to you should serve as the Greenwich scale furthermore according to this source https://research.collegeboard.org/programs/sat/data/archived/cb-seniors-2010/tables they have a 92% participation rate which explains why their sat mean is the lowest, 467 for Math, 468 for CR and 454 for Writing, now given they are not full anglo white we shouldn’t assume that this SAT mean is necessarily 100 but they do have a participation rate of 92% so it ends up evening it out, assuming it is 100 or thereabouts we can now have a much better grasp of what for example a 650 on the Sat math is in comparison to the general white population, I couldn’t immediately find the SD of the math section that year I’m too lazy to dig it up 😂 but assuming it is 113 :
[(650- 467)/113]*15 + 100 = 124.43 Lets make it 125, which is pretty congruent with your inferences, just thought I could use my tyro understanding of basic statistics to show it is possible to convert sat scores into Iq or at least have a sense of where you might be in the bell curve even if you’re not as knowledgeable as pp or other great HBD thinkers.