Disclaimer
This data was collected using the Kids Labs mobile app from 2012-2-4 to 2012-2-14. During that 10-day period, we collected data from 1,772 users. Every user provided the birthday and gender of their child and permission to use their data in publications. This article will summarize the results and provide an early look into the data from this ongoing experiment.
Tracing
The app randomly selects a shape for the user to trace. When the user finishes tracing the shape, the time and average distance from the perfect line is reported to a server.
Unsurprisingly, the line is the simplest shape to trace, lowest time and deviation. However, there are several interesting results in this data. The star shape takes the longest time, but deviation is average. Something about tracing a star slows down the user and they really focus on tracing it perfectly. The circle shape has the largest deviation, but a fast time, perhaps showing an unjustified level of confidence. Drawing a perfectly round circle is a common training exercise for artists, and given the deviation results shown above that seems like an appropriate choice.
Categories
In this test, we show the user 6 objects, 2 sets of 3 matching objects. The objects either share the same color, shape or nothing. The color was randomly selected from one of these four colors: green (#5CC151), purple (#D40072), yellow (#FEE000), cyan (#00AAD2). The shape was randomly selected from: triangle, square, pentagon, hexagon, octagon, circle, star.
Results:
| Color | Shape | Time |
| Different | Different | 12.77 |
| Same | Different | 13.59 |
| Different | Same | 14.20 |
When the color & shape are different this provides our control data. We can compare the other results with this control to determine whether color or shape is more significant. In our data set, shape was more significant than color, specifically 43% more important. This is an important finding for anyone developing apps for children, use shapes to differentiate instead of color.
Age
Interestingly, many parents decided to create an account for themselves, so I have data for users over 40. However, to keep this analysis accurate I limited my results to ages 2 to 11, which had enough samples (>200 per year) to draw valid conclusions.
Let’s jump right into the data:
So, what did we learn from this data. First off, many parents cannot help the urge to finish the exercise when their child fails. For tracing and categories, we can see that the results are worse at age 3 than 2, which is highly unlikely. We can conclude that a large number of 2 year old children cannot complete the tracing & category exercises and their parent finished it for them. Looking at the results for Clock Time, it shows the same pattern but all the way until age 6. The clock exercise asks the child to move the hour & minute hand of clock to match the goal time, which is apparently too difficult for children under age 6. Personally, I have a 4 year old daughter and she has never finished this exercise, so this is not a surprising finding. If we ignore the parents cheating, we can see some interesting data about improvements in skills over time. Tracing time barely changes, but deviation improves from 29.48 pixels to 17.30 pixels or (7/32” to 1/8” or 56mm to 33mm). That is 58% from age 3 to age 6, then from age 6 to 42 there is no statistically significant improvement. Age 6, is also the age when the children can complete the clock exercise. It is impossible this shows that the clock exercise requires achieving a certain level of hand eye coordination. It is surprisingly, that there is no improvement beyond the age of 6. I wonder if it is a limitation of the iPad device and its ability to precisely measure finger position. This is also great data for anyone developing apps for children, make sure your click targets are 1.7 times large than what you would use for an adult.
Gender
Prepare yourself for disappointment. I could find no statistically significant differences between male & female performance on these exercises.