Skip to main content
. 2018 May 21;28(10):1561–1569.e3. doi: 10.1016/j.cub.2018.03.069

Figure 1.

Figure 1

Option-Generation Task

(A) The option-generation task required participants to draw, in 4 min, as many different paths as possible from the start point (bottom circle) to the end goal (top circle). Paths appeared as participants drew them and remained displayed on the touchscreen during the task so that participants did not have to remember them.

(B) To quantify uniqueness, each path was first parameterized by 200 equally spaced points along its length. The “difference” between any two paths is then taken as the sum of the distance between corresponding points. This distance metric also includes the first and second derivatives in order to better account for curvatures in trajectories. Uniqueness of a path is then defined as the “distance” between it and the most similar path generated (by any participant in the three studies of this paper). Distances between a participant’s paths were also projected into a 2-dimensional subspace using multidimensional scaling to visualize how individuals explored the space of possible paths by treating each generated path as a point. Points that are closer together indicate more similar paths and vice versa.

(C) Illustrations of the trajectories of the 69 paths generated by one participant (left), his corresponding points in 2-dimensional similarity subspace after multidimensional scaling (middle), and the pairwise distance matrix (right).

See also Figure S1.