The spreadsheet contains the data presented in Figure 1 in the paper. Values in each row are from an individual subject. Each datum is a mean proportion of the reported dominance of a stimulus type for each condition. The different conditions were the dichoptic stimulus pairings, including the single colour controls: RR (Red-Red), GG (Green-Green), BB (Blue-Blue), YY (Yellowish/Orange-Yellowish/Orange) and the dichoptic colour pairings: RG (Red-Green), BY (Blue-Yellowish/Orange), RB (Red-Blue) and GY (Green-Yellowish/Orange). P(CCW) = Mean proportion (of the total viewing time) where counter-clockwise rotation was reported as dominant. P(CW) = Mean proportion (of the total viewing time) where clockwise rotation was reported as dominant. P(CCW) MB = Mean proportion of counter-clockwise rotation dominance where colour misbinding was also reported (expressed as a proportion of the total viewing time). P(CW) MB = Mean proportion of clockwise rotation dominance where colour misbinding was also reported (expressed as a proportion of the total viewing time). The remaining columns simply sum the data across direction of rotation, i.e. CCW+CW for both the raw direction dominance and the misbinding. It was these summed proportions that were arcsine-transformed and analysed with paired t-tests (or one-sample t-tests in the case of the non-opponent colour pairings) as reported in the paper. Pilot data for an earlier version of this experiment (presented in the electronic supplementary material) is not included in this file. Ryan T. Maloney ryan.maloney@sydney.edu.au