## Comparing the chance probability and the observed probability of getting both length and orientation correct within response types (for Group 1).

In each panel the y-value of a data-point is the proportion of trials, for a given participant responding to a given combination of length and orientation, where both length and orientation were correct. Therefore, in total, there are 140 data-points in each panel (seven participants performing each of the 20 combinations of length and orientation). Each data-point is based on 16 trials. The x-value of a data-point is derived from a model of independence. It is calculated by multiplying the marginal probability of getting a specific length correct (across all orientations) with the marginal probability of getting a specific orientation correct (across all lengths). The specific length and orientation is chosen so that it corresponds to the combination of length and orientation used to derive the y-value. The red line is y = x and the grey line is the best linear fit. The lines show how closely matched the observed probability is to the chance probability. For motoric responses in Group 1 the slope of the grey line was 1.023 (and intersected the y-axis at 0.000). For symbolic responses in Group 1 the slope of the grey line was 1.055 (and intersected the y-axis at 0.0032). For motoric responses in Group 2 the slope of the grey line was 1.011 (and intersected the y-axis at 0.01). For symbolic responses in Group 2 the slope of the grey line was 1.096 (and intersected the y-axis at 0.0252).