Monkey see, monkey do?
Try your hand at imitation

Here is an chance to test how accurately you can imitate a simple action that you have just seen someone else perform. The models, which you can view and then try to imitate, are based loosely on ones used by Sekuler, Siddiqui, Goyal and Rajan (2003) "Reproduction of seen actions: Stimulus-selective learning," Perception, 32 839-854 and by Agam, Bullock and Sekuler (2005) "Imitating unfamiliar sequences of connected linear motions," Journal of Neurophysiology.

The to-be-imitated models comprise an open hand, which moves through a series of quasi-random, linked segments. That set of segmented movements defines the action you try to reproduce, using the mouse and keyboard on your computer. Stimulus models (below) vary in complexity. The ones labelled "Practice" are easy, but as the number of segments (components) increases, so too does the difficulty of reproducing what you've just seen.

There is a blue bar at the bottom of this page. The demonstrations require that your browser window is large enough to make that blue bar visible.

Here's how to perform your imitation (Don't worry: these instructions are repeated after each stimulus model). Once the hand has finished generating the model stimulus, a response screen will appear. Start by clicking on the screen's white disk. This will cause a blue trail to track any subsequent movements of the mouse. To reproduce the movements made by the stimulus hand, move the mouse along the segmented path you remember the hand having taken. Each time you reach the end of a segment, click the mouse. When you finish the last segment of your imitation, press the Space Bar. To compare your imitation to the hand's actual path, click the button in the lower right hand corner.

If you dare to check how accurately you reproduced what you saw, you can click a button that will superimpose your imitation on the actual stimulus model. (Sekuler, Siddiqui, Goyal and Rajan used a complex quantitative measure of accuracy; that method has not implemented here.)

As you proceed through the stimulus models, you will discover that your reproduction's fidelity declines as the number of components in the model grows. Also, if you wait a while before starting your imitation, you may find that your memory, and hence your accuracy, declines. Finally, if you get several people to tackle the task, you may discover considerable differences from one person to another in performance.