According to IQ Comparison Site, an SAT score of 1485/1600 corresponds to an IQ of about 144. According to Ivy West, an ACT of 33 corresponds to an SAT of 1470 (and thence to IQ of 143).
Only if you took the SAT before 1994. Here’s the percentiles for SATs taken in 2012; someone who was 97th percentile would get ~760 on math and ~730 on critical reading, adding up to 1490 (leaving alone the writing section to keep it within 1600), and 97th percentile corresponds to an IQ of 128.
Here’s a classic calibration chart.
An important part of the calibration chart (for people) is the frequency of times that they provide various calibrations. Looking at your table, I would focus on the large frequency between 10% and 30%.
I’ll also point out that fixed windows are a pretty bad way to do elicitation. I tend to come from the calibration question from the practical side: how do we get useful probabilities out of subject-matter experts without those people being experts at calibration? Adopting those strategies seems more useful than making people experts at calibration.
Only if you took the SAT before 1994. Here’s the percentiles for SATs taken in 2012; someone who was 97th percentile would get ~760 on math and ~730 on critical reading, adding up to 1490 (leaving alone the writing section to keep it within 1600), and 97th percentile corresponds to an IQ of 128.
An important part of the calibration chart (for people) is the frequency of times that they provide various calibrations. Looking at your table, I would focus on the large frequency between 10% and 30%.
I’ll also point out that fixed windows are a pretty bad way to do elicitation. I tend to come from the calibration question from the practical side: how do we get useful probabilities out of subject-matter experts without those people being experts at calibration? Adopting those strategies seems more useful than making people experts at calibration.