The Draw-A-Person Test

Imagine a world where intelligence is measured like this:

A child sits down at a desk. She is given a piece of paper and a crayon. Then, she is asked to draw a picture of a boy or girl. “Do the best that you can,” she is told. “Make sure that you draw all of him or her.” If the child hesitates, or asks for help, she is gently encouraged: “You draw it all on your own, and I’ll watch you. Draw the picture any way you like, just do the best picture you can.”

When the child is done drawing, the picture is scored. It’s a simple process, with little ambiguity. One point is awarded for the “presence and correct quantity” of various body parts, such as head, eyes, mouth, ears, arms and feet. (Clothing gets another point.) The prettiness of the picture is irrelevant. Here are six drawings from four-year olds:

The Draw-A-Person test was originally developed by Florence Goodenough, a psychologist at the University of Minnesota. Based on her work with Lewis Terman – she helped revise and validate the Stanford-Binet I.Q. test – Goodenough became interested in coming up with a new measure of intelligence that could be given to younger children. And so, in 1926, she published a short book called The Measurement of Intelligence by Drawings which described the Draw-A-Person test.* Although the test only takes a few minutes, Goodenough argued that it provided a window into the child mind, and that “the nature and content of children’s drawings are dependent primarily upon intellectual development.” In other words, those scrawls and scribbles were not meaningless marks. Rather, they reflected something fundamental about the ways in which we made sense of the world. The act of expression was an act of intelligence, and should be treated as such.

In her book, Goodenough described the obvious benefits of her intelligence test. It was fast, cheap and fun. What’s more, it seemed to be measuring something real, as children tended to generate a consistent set of scores over time. (In other words, the test was reliable.) And yet, despite these advantages, the Draw-A-Person test largely fell out of favor by the 1970s. One explanation is that it was lumped in with other “projective” techniques, such as the Rorschach Test, that were repeatedly shown to be inaccurate, too tangled up with psychoanalytic speculation.

However, a new study by Rosalind Arden and colleagues at King’s College London suggests that Goodenough’s test still has its uses, and that it manages to quantify something important about the developing mind in less than ten minutes. “Goodenough’s genius was to take a common childhood product and see its potential as an indicator of cognitive ability,” they write. “Our data show that the capacity to realize on paper the salient features of a person, in a schema, is an intelligent behavior at age 4. Performance of this drawing task relies on various cognitive, motoric, perceptual, attentional, and motivational capacities.”

How’d the scientists show this? By giving the test to 7,752 pairs of British twins, the scientists were able to compare the drawing performance of identical twins, who share all of their genetic material, with that of non-identical twins, who only share about half. This allowed them to tease out the relative importance of genetics in determining scores on the Draw-A-Person test. (All of the twin pairs were raised in the same household, at least until age 4, so they presumably had a similar home environment.) The results were interesting, as the drawings of identical twins were much more similar than those of non-identical twins. There is no drawing gene, of course, but this result does suggest that the sketches of little kids are shaped by their genetic inheritance. In fact, the results from a single drawing were as heritable among the twin pairs as their scores on more traditional intelligence tests.

Furthermore, because the researchers had scores from these intelligence tests they were able to compare performance on the Draw-A-Person test with a subject’s g factor, or general intelligence. The correlations were statistically significant but relatively modest, which is in line with previous studies. This means that one shouldn’t try to predict IQ scores based on the scribbles of a toddler; the two variables are related, but in weak ways.

However, a more interesting result emerged over time, as the scientists looked at the relationship between drawing scores at the age of 4 and measures of intelligence a decade later, when the twins were 14. According to the data, the children’s pictures were just as predictive of their intelligence scores at the age of 14 as various intelligence tests given at the age of 4. "This study does not explain artistic talent,” write the scientists. “But our results do show that whatever conflicting theories adults have about the value of verisimilitude in early figure drawing, children who express it to a greater extent are somewhat brighter than those who do not." 

Such studies trigger a predictable reaction in parents. I've got a three-year old daughter - I couldn't help but inspect her latest drawings, counting up the body parts. (There's even an app that will help you make an assessment.) But it's important to note that this is all nonsense; the science does not support my anxieties. "I too fossicked around in old drawers to look for body-parts among the fridge-magnet scrawls of my former 4-year old," Dr. Arden wrote in an email. "I realised quickly the key question was not 'is she bright?', but 'did we have fun? Did I treasure that wonderful, lightspeed flashing childhood properly?'" In a recent article put out by King's College, Arden expands on this idea, observing that while her "findings are interesting, it does not mean that parents should worry if their child draws badly. Drawing ability does not determine intelligence, there are countless factors, both genetic and environmental, which affect intelligence in later life.”

I find this study most interesting as a history-of-science counter factual, a reminder that there are countless ways to measure human intelligence, whatever that is. We've settled on a particular concept of intelligence defined by a short list of measurable mental talents. (Modern IQ tests tend to focus on abilities such as mental control, processing speed and quantitative reasoning.) But Goodenough’s tool is proof that the mystery of smarts has no single solution. The IQ test could have been a drawing test.

This sounds like a silly conjecture. But it shouldn’t. As the scientists note, figurative art is an ancient skill. Before there were written alphabets, or counting systems, humans were drawing on the walls of caves. (There’s evidence that children participated in these rituals as well, dragging their tiny fingers through the wet clay and soft cave walls.) "This long history endows the drawing test with ecological validity and relevance to an extent that is unusual in psychometrics," write the scientists. After all, the Make-A-Person test measures one of the most uniquely human talents there is: the ability to express the mind on the page, to re-describe the world until life becomes art, or at least a crayon stick figure.

*Goodenough originally called it the Draw-A-Man test, but later realized that the gendered description made it harder for young girls.

Arden, Rosalind, et al. "Genes Influence Young Children’s Human Figure Drawings and Their Association With Intelligence a Decade Later." Psychological Science (2014)