A diagram of the experimental setups for the four experiments included in the study is reported in
Figures 1A–D. All experiments shared the same logic and design, which we describe first, followed by the specific features of each individual experiment.
For all experiments, participants took part in two experimental sessions, at least 24 hours apart. Each session was divided into two trials of 3 minutes each (6 minutes in total), except for experiment 2, where four 3-minute-long trials (12 minutes in total) were tested. Participants viewed the monitor from a 57 cm distance; a chin and forehead rest stabilized head position. The stimuli consisted of small circular gratings presented dichoptically in central view; they were inscribed in a binocular frame to facilitate fusion. Participants reported rivalrous alternations through the computer keyboard, by continuously pressing one of three keys to report exclusive percepts of orthogonally oriented gratings (right arrow for clockwise and left arrow for counterclockwise) or a mixture of those (piecemeal or fusion: down arrow key). The orientation of gratings presented in either eye was counterbalanced across participants and switched on every trial to avoid adaptation. For experiment 1, the swapping procedure was done every 90 seconds, that is, halfway through a trial.
Experiments mainly differed in the method used for dichoptic stimulation. Experiments 1 and 2 used a mirror stereoscope placed in front of an LCD monitor (BenQ XL2420Z, 1920 × 1080 pixels, 144 Hz refresh rate, Taipei, Taiwan). Experiment 3 and the first session of experiment 4 used CRS ferromagnetic shutter goggles (Cambridge Research Systems, Kent, UK) and a CRT monitor (Barco 6551, 800 × 600 pixels, 140 Hz refresh rate, Kortrijk, Belgium). The second session of experiment 4 used anaglyph red-blue goggles and a LED monitor (LG IPS 24EA53, 1920 × 1080, 60 Hz refresh rate, Seoul, South Korea).
There were also differences in the stimuli used to induce rivalry (recall that these experiments were run for independent studies). In experiments 1 and 3 and in the first session of experiment 4, stimuli were monochromatic sinusoidal gratings (orientation: ±45°, spatial frequency: 2 cpd, contrast: 50%, size: 3° or 2° in experiments 1 and 3 and experiment 4, respectively) presented against a uniform gray background (experiment 1: luminance 110 cd/m2, C.I.E. x = 0.305, y = 0.332; experiment 3 and 4: luminance 37.4 cd/m2, C.I.E. x = 0.442, y = 0.537). In the second session of experiment 4, stimuli were red and blue gratings (orientation: ±45°, size: 3°, spatial frequency: 2 cpd, maximum luminance: 0.5 cd/m2), presented against a black uniform background. In experiment 2, stimuli were 3° disks, one white (maximum screen luminance 295 cd/m2) and one black (minimum screen luminance 10 cd/m2) shown against a uniform gray background (luminance 152 cd/m2). To discourage binocular fusion, the disks were overlaid with orthogonal gray lines (45° clockwise or counterclockwise, 0.033° or 1 pixel wide, and 0.5° apart).