Fig. 1. Mean RT for Experiment 1 for Mir and Non groups, for the three experimental conditions.

Fig. 2. Mean RT for Experiment 2 for Mir and Non groups, for the three experimental conditions.

Fig. 3. Schematic of the apparatus and setup for all experiments (A), the four possible diagrams that the model sees to instruct her what movement to produce on each trial (B), and an example of the templates shown in the right target marked condition (C). See text for details.

Cumulative errors for the (H)and, (T)arget, and (B)oth hand and target, as a function of task condition and Mir and Non modes of imitation on each experiment

*In each case, 1 subject switched modes during the block and produced in excess of 45 errors. All other subjects produced 0 errors.

**In each case, 2 subjects switched modes during the block and produced in excess of 42 errors each. All other subjects produced 0 errors.

***In each case, 1 subject switched modes during the block and produced in excess of 43 errors. All other subjects produced 0 errors. Note that in the other conditions with high error counts, the sum of errors is shown across all subjects, and approximately 6 subjects in total made up the cumulative total.

Mean RT (standard error) for each condition in Experiment 3

Mir = mirror mode, Non = Non-mirror mode, H = hand, T = target, Dif = difference score (Non-Mir).

Number of participants in Experiment 4 who performed in each mode for each condition and (total error count), followed by number of participants who shifted modes in each condition compared to the initial baseline condition

Base 1 = initial baseline condition, Base 2 = second baseline condition (block 4), Mir = mirror mode, Non = non-mirror mode, T = target, H = hand. Total error count reflects the cumulative error across 72-trial blocks for all participants combined.