In his excellent book A Question of Intelligence: The IQ Debate in America, former Fortune magazine editor Daniel Seligman describes what it’s like to take the WAIS-R IQ test.  One page 8, he describes taking the Digit-Symbol subtest.  According to the school board I attended as a kid, this test measures:

Rapid eye-hand coordination
Grapho-motor speed
Visual fix
Immediate visual memory
Motor control

Seligman writes:

Marching toward the end of the test, we now reached Digit Symbol, a performance test that involved some more memory skills.  What you have to remember this time were visual symbols, so I approached the test with minimal confidence.  You begin by inspecting a table that shows nine simple symbols–a circle, a plus sign, etc–and are told that each represents a digit from 1 through 9.  Then you are shown several rows of digits, arranged in random order.  The object is to draw the appropriate visual symbols in a box under each digit and to complete as many boxes as you can in the allotted ninety seconds.  If you can memorize the symbols as you are doing the exercise, you will of course go much faster, and complete more boxes, than if you had to keep referring back to it; at least I knew I would make a fair number of mistakes if I tried to rely on memory.  It struck me midway through Digit-Symbol, that it would have been nice to know something about the scoring tradeoffs between accuracy and speed.  Possibly I was over concerned about accuracy;  anyway, I ended up with a so-so total of fifty-four symbols.  The subtest is designed to measure attentiveness and quickness, and, not surprisingly, it is tough for us old folks.  Of all the subtests (I later discovered), it shows the greatest age-based decline in raw scores.

According to the WAIS-R manual, Seligman’s raw score of 54/93 equated to a scaled score of 9 in the peak age group (20-34) and a scaled score of 12 in Seligman’s age group (55-64).

But because WAIS-R norms were a decade old when Seligman was tested, and the Flynn effect increased WAIS Digit Symbol scaled scores by 0.71 points a decade from 1978 to 1995 (Flynn, 2012), his scaled scores must be reduced to 8.22 and 11.29 respectively.  An age adjusted scaled score of 11.29 is equivalent to an above average IQ of 106 (U.S. norms; 104 U.S. white norms) on that subtest.

When I was in junior-high school (I’m now in my 30s), all three of my best friends had overall IQs above 100, yet all three were way below average at Digit-Symbol.  While I consider this a poor test of intelligence generally, I think people who score low on it have something wrong their brains that prevents them adapting.  In the sixth-grade I overheard our teacher say of one of my digit-symbol impaired friends that no matter how hard he tried, he was never going to get anywhere in this World.  I strongly suspect another Digit-Symbol impaired friend robbed my house when I was in high school to support a drug habit.  My third Digit-Symbol impaired friend got an advanced degree and a job in the government, but got laid off and last I heard, works in a low status service job at a retail store.  I call it Digit-Symbol destiny:  the curse of Digit-Symbol.

The great David Wechsler talks about a subject who got an IQ equivalent of 50 on Digit-Symbol and was so deviant, that during the testing session, he asked the examiner to have sexual intercourse with him!

David Wechsler was a Digit-Symbol enthusiast, writing in The Measurement and Appraisal of Adult Intelligence (1958):

The Digit Symbol or Substitution Test is one of the oldest and best established of all psychological tests. It is to be found in a large variety of intelligence scales, and its wide popularity is fully merited…The question that remains is whether speed as well as power should be given weight in the evaluation of intelligence. The author’s point of view is that it should…Neurotic and unstable individuals also tend to do rather poorly on the Digit Symbol (as indeed on all other substitution tests). The inferiority of neurotic subjects on tests of this kind was noted as long ago as 1923, by Tendler. Tendler suggested that this was due to some sort of associative inflexibility in the subject, and a tendency toward mental confusion. More obviously neurotic subjects do badly on this test because they have difficulty in concentrating and applying themselves for any length of time and because of their emotional reactivity to any task requiring persistent effort. The poor performance of the neurotic represents a lessened mental efficiency rather than an impairment of intellectual ability.

More recent research suggests that autistics tend to do poorly on this test.  Too bad autism wasn’t talked about in Wechsler’s day; he would have had a field day with it.  The poor performance of autistics might be because the test perhaps measures executive function, or maybe it’s because autistics are obsessively concerned with accuracy which slows them down on this task.