What kind of copying is it?

Two questions carve the family of imitative behavior. Did the observer copy the demonstrator’s action form (the specific means)? And did they reproduce the goal (the end state)? Toggle each and watch the classification. The four cells are the categories shown in Figure 1.

Reproduced the goal?noyesgoal: nogoal: yesCopied the form?Mimicryform yesTrue imitationform yesNo copyingform noEmulationform no
True imitation. The observer reproduced the demonstrator's specific action form and brought about the same end. Copying the means is the defining mark of imitation proper.

The compatibility (congruency) effect

In an automatic-imitation task the observed movement is irrelevant, yet a matching movement speeds the response and a mismatching one slows it. The compatible mean is fixed at 452 ms; set the incompatible mean, the standard deviation of the difference, and the sample size, and watch the congruency effect, Cohen’s d, and the paired t. The defaults reproduce the article’s worked example. Values are illustrative and computed locally.

Incompatible RT487 ms
SD of difference70 ms
Participants (n)24
Compatible452 msIncompatible487 ms35 ms
Congruency = 487 − 452 = 35 ms. Cohen’s d = 35 / 70 = 0.50 (medium effect). With n = 24, SE = 70/√24 = 14.29 ms, so t(23) = 2.45, conventionally significant.

Over-imitation and the transparent box

A demonstrator opens a puzzle box with two actions: one that does nothing (tapping the top) and one that releases the latch. When the box is opaque, the causal structure is hidden; when it is transparent, the pointlessness of the first action is plain to see. Choose the learner and the box, and compare how often the causally irrelevant step is copied. Rates are illustrative of the Horner & Whiten pattern.

Irrelevantstep copied75%50%
Over-imitation. The child keeps copying the useless step at 75% even though the transparent box makes its irrelevance obvious. High-fidelity copying of the means overrides efficiency.