Figure 1
Paradigm and behavioral division with the effect of goals. (A) The four task stages of each trial in the memory task performed by the participants (adapted from Chen et al. (Current Biology, 2018)). (B) An aerial view of the circular arena. The drop error is defined as the relative distance between the subjective response location and the reallocation. (C) Two types of groups are defined by drop error in behavior. The top row is GoodPerf, and the bottom row is BadPerf. The middle column is the trajectory of the participants, achieving extensive coverage in both groups. The right-most column shows the movement direction of the participants, which is evenly distributed at all angles between the two groups (Rayleigh’s tests for non-uniformity, p > 0.05). (D) Comparison of experimental data between GoodPerf and BadPerf. A significant difference was not observed between the two groups (t test, p > 0.5). (E) Comparison of performance in the trials corresponding to GoodPerf and BadPerf goal objects. The difference in drop error between the two groups was significant (t test, p < 0.001). (F) An example of the time sequence of GoodPref and BadPref trials during the whole experiment period (taking subject #1 as an example), and the trial sequence number index of each type group is calculated based on the average sequence number of all trials in this category. (G) Temporal distribution of the two groups in the experiment. We define the trial serial number index to represent the temporal distribution, represented by the average serial number of all trials in each group in the experiment. Statistics showed that Goodpref and Badpref were evenly distributed in the experiment, without significant difference (paired t test, p < 0.05).
