When I was a child, the teachers couldn’t figure out why I never paid any attention in class. After a couple of visits to a psychologist, and following the administration of an IQ test on which I scored 130, he confirmed I my mind was simply in 5th gear as the teacher was presenting second grade material.
In the Spring of 1985, a bunch of people were gathered on the lawn of my college apartment complex for a picnic. I recognized a friend and stepped outside to join her for some conversation. She said her father, one of my former professors, was holding a Mensa outreach picnic and invited me to join the group. A couple of weeks later, I took the test, and walla! I was in Mensa.
I have since taken a number of IQ tests such as the one below, with scores of 120, 133, 139, 140, 167, and 180+. What’s going on? Why the variation in scores? Let’s take a look at an IQ test I took this morning as we examine why.
Even though I answered all questions correctly, the test shows a lower IQ than my actual IQ. Nevertheless, it’s statistically accurate, in that there were only 20 questions. Even though I scored a 20/20, provided the IQ test contains questions of equally distributed difficulty, one cannot say for statistical certainty that it’s any higher than 133. There simply aren’t enough questions. A score of 40/40 might result in an IQ score of 140; 60/60 might result in an IQ score of 150, and 100/100 questions might result in an IQ score of 160. The more questions on the test, along with the more people tested using that particular test, then the finer the granularity with which the test can pinpoint your score. When it comes to scores with the middle 85%, it’s probably reasonable accurate, +/- 10 to 15 points. When it comes outliers, however, it cannot differentiate between a score of perhaps 125 and a 200, so it picks the statistical median of all scores higher than 125, which is 133. A test with more questions would be able to pinpoint that to perhaps 128, or 132 or 150+.
The thing about it is the furthermore down or up the scale you fall, the more of an outlier of your score, then the more questions are required to definitely assess your score AND the more people have to take the test against which your performance can be measured.
With that in mind, let’s see where these scores originated:
120: Given to me by a forensic psychologist at my request, along with a large battery of other tests, in order to determine suitability for visitation and custody during divorce proceedings. I’d experienced severe chronic insomnia for six months straight prior to the test, and probably got all of about four hours of sleep the night before, so, no, I didn’t do very well!
133: Score obtained by correctly answering 20 out of 20 questions. You can take the test yourself by following the links, above.
139: Test I took in second grade, administered by a child psychologist. The test didn’t go any higher than 139, so 139 it was.
140: Test I took in the Air Force, as part of a low-key study conducted by our squadron. Again, the test didn’t go any higher.
167: An experimental but reasonably accurate test designed to measure people with a high IQ.
180: An actual IQ test. The score doesn’t go any higher. I think I was lucky, as I was feeling exceptionally well-rested and alert that day!
So, what’s my actual IQ? Well, it’s certainly higher than 120, that’s for certain. I believe it’s higher than 139/140, as those were my scores — and the top score — on two standard IQ tests widely used and accepted throughout the arena of psychology. I’m reasonably sure it’s higher than 150, as I did become a member of Mensa. I think it may be as high as 180 on my best days, when I’m firing on all 100 billion neurons.
However, given more than half a century of experience interacting with other people, I’m inclined to believe I have an IQ of 167 +/-15 pts. Again, while it’s relatively easy to pinpoint an IQ ranging in the middle 83% of the pack, between 80 and 120, the further out you go in either direction, the more difficult it is to accurately determine one’s IQ score.
If you want to take either a short or a long free online Stanford-Binet IQ test, click here. For kicks and grins, just to see what a really low score might feel like, I took the long test of 100 questions, randomly answering at least 3 out of 4 questions, and only thinking about the ones where I liked the way the question looked. If it looked like I might have to think about it, I skipped it. My score put me in the 83rd percentile, which corresponds to an IQ of about 116.