Measuring children's associative divergent ability through drawings and semantic similarity

Imagine your brain is a bustling train station. Every time you see something—a dot, a word, a picture—trains of thought shoot out in different directions. Some trains stay on the main tracks, heading to obvious places like "That's a circle!" Others zoom off into wild, scenic routes, turning that circle into a planet, a pizza, or a giant eye. This ability to send your thoughts racing down many different tracks at once is called divergent thinking. It's the spark behind creativity, the "what if" engine that helps us solve problems and invent new things.

For a long time, scientists trying to measure this spark had a tricky job. The standard way was to ask people to fill out surveys or answer questions, which is like trying to measure the speed of a race car by asking the driver how fast they felt they were going. It's subjective and hard to compare. Recently, a new idea has emerged: instead of just words, let's look at drawings. Since kids are natural artists, their sketches might reveal how their minds wander. But how do you turn a scribble into a number? That's where modern technology steps in, using "smart computers" (AI) to read the drawings and translate them into language, allowing us to measure exactly how far apart those trains of thought are.

This is the story of a new study that tried to map the creative minds of 827 children in Beijing. The researchers wanted to see if they could measure a child's "associative divergent ability"—basically, how wildly and uniquely they can connect ideas—by watching them turn a single black dot into a full picture. They didn't just ask the kids to draw; they used a clever mix of AI and math to turn those drawings into data, revealing surprising patterns about how children's imaginations work, how they change as they get older, and how a tiny hint can either open a door or slam it shut.

The Experiment: A Dot, A Dream, and a Smart Computer

The researchers set up a simple but powerful game called "Small black dot, please change!" They handed out blank sheets of paper with a single black dot in the middle and asked the students to use their imagination to complete the drawing. To make things even more interesting, they added "gradient tasks." Imagine if instead of just a dot, the starting point was a specific object like a brain, a magnet, a plant, or a pair of glasses. Sometimes, they even added a second item to the starting picture (like a ruler next to a brain) to see if that extra clue would force the children's minds to focus more narrowly.

Once the 827 drawings were finished, the researchers didn't just look at them with human eyes. They fed the images into a large language model (a super-smart AI) that acted like a translator. The AI looked at each drawing and described what it saw in words, categorizing it into themes like "Nature," "Sports," "Magic," or "Science." Then, the researchers used a mathematical tool called semantic similarity to measure how close or far apart these descriptions were. Think of it like a map: if two kids draw a sun and a tree, their ideas are close together on the map. If one draws a sun and the other draws a spaceship, their ideas are far apart. The goal was to see how "spread out" the children's ideas were.

What They Found: Nature Rules, and Age Changes the Map

The results painted a vivid picture of the children's creative landscape. First, the most popular destination for these wandering minds was Nature. Whether it was a dot or a magnet, the kids most often turned their starting point into scenes of the sun, clouds, trees, or animals. In fact, the distribution of their ideas followed a "power-law," meaning a few themes were super popular, while many others were rare.

When it came to how "spread out" the ideas were (the semantic similarity), most children's drawings landed in a specific range, suggesting a common "creative zone." However, the starting point mattered a lot. When the researchers gave the kids a specific object to start with (like a magnet), the ideas became more convergent—meaning the children's thoughts were closer together, sticking to the theme of that object. It was as if the magnet pulled their imaginations toward a specific magnetic field. But here's the twist: not all hints worked the same way. A pair of glasses with a microscope attached made the ideas very narrow, but a brain with a ruler attached actually made the ideas more spread out. Why? The researchers suspect that some kids didn't understand the connection between the brain and the ruler, so they ignored the hint and let their minds wander freely!

Growing Up and Gender Differences

The study also looked at how age and gender affected creativity. Surprisingly, gender didn't matter at all. Boys and girls showed the exact same patterns in how they diverged from the starting dot.

Age, however, told a different story. As children got older, their drawings changed significantly. Younger kids (grades 3 and 4) loved drawing animals and nature scenes. But as they moved up to grades 5 and 6, they drew far fewer animals and nature scenes, and instead, they started drawing people and portraits much more often. The researchers found a link between this shift and what the children were learning in school. As their science textbooks focused less on biology and more on other topics, the children's drawings followed suit. It seems that what we learn in class directly shapes the tracks our creative trains take.

The "Aha!" Moment: When Does the Divergence Stop?

One of the most fascinating questions the study tackled was: When does the creative process actually end? Does a child stop thinking creatively the moment they decide, "Okay, this dot is an eye," or do they keep diverging even after that decision?

To find out, the researchers looked at the top three things kids thought the dot was: an eye, a head, or the center of a flower. They compared the creativity of kids who stopped there versus those who kept going. They found that if a child saw the dot as a flower center, their ideas became very narrow and focused on plants. But if they saw it as an eye or a head, they had a huge amount of space left to explore! This suggests that the creative process doesn't stop when you identify the object; it continues as you build the rest of the picture. The "eye" and "head" categories left the door wide open for more wild ideas, while the "flower" category closed the door tight.

The Takeaway

This study suggests that we can measure children's creativity not just by asking them questions, but by watching how they draw and using smart computers to analyze the "distance" between their ideas. It shows that while children's imaginations are naturally wild and often drawn to nature, their learning environment and the specific hints they are given can steer their thoughts in very specific directions. Most importantly, it proves that creativity isn't a single flash of inspiration; it's a journey that keeps going, even after the first idea is formed. By understanding these patterns, we might be able to design better ways to help children keep their creative trains running on the most exciting tracks possible.