Figure 1
Experience curves as a function of trial number, where Trial 1 is the first IT stimulus presented, and Trial 80 is the last IT stimulus presented. Left = Response accuracy (i.e., percentage of responding participants who provided a correct response), and percentage of 147 participants who provided a response within the allocated time. Right = Mean response time of the participants who provided a response, together with 10th and 90th percentiles. Exponential fits, y = 1/(a + b*exp(–c*x)), are shown where x is the trial number, and a, b, and c are fitted parameters. For the ‘response accuracy’ curve, a = 0.0131, b = 0.00387, c = 0.155 (r2 = 0.53). For the ‘response provided’ curve, a = 0.0105, b = 0.0198, c = 1.033 (r2 = 0.87). For the ‘mean response time’ curve, a = 0.00100, b = –0.000516, c = 0.0811 (r2 = 0.94). Note that the IT stimuli were presented in random order.
Figure 2
Response accuracy (i.e., percentage of responses that were correct), percentage of trials with a response within the allocated time, and mean response time as a function of exposure time of the Pi-figure. The means and standard deviations are provided in Table S1.
Table 1
Inspection time task performance per self-reported illusion (N = 147).
| Response accuracy (% of trials in which the participant responded) | No response (% of all 80 trials) | Mean response time (ms) | |
|---|---|---|---|
| Mean (SD) | Mean (SD) | Mean (SD) | |
| Growing illusion (n = 56) | 74.52 (14.23) | 3.98 (5.74) | 1141 (522) |
| Brightness illusion (n = 17) | 80.33 (7.76) | 0.82 (1.01) | 860 (273) |
| No illusion (n = 74) | 72.33 (16.87) | 6.22 (12.81) | 1202 (560) |
| Welch’s test | Welch’s test | Welch’s test | |
| Growing vs. no illusion | t(126.4) = 0.80, p = 0.426 | t(106.8) = 1.33, p = 0.185 | t(122.4) = 0.64, p = 0.523 |
| Brightness vs. no illusion | t(55.3) = 2.94, p = 0.005 | t(76.8) = 3.57, p < 0.001 | t(51.3) = 3.68, p < 0.001 |
Figure 3
Percentage of trials in which participants were blinking, for each time sample. A distinction is made between trials where participants provided a correct response (n = 8308) and trials where participants provided an incorrect response (n = 2736). Top figure: results time-locked to the stimulus (occurring at t = 1.1 s). Vertical lines are shown for the moment the fixation marker (X) disappeared, the moment the Pi-figure was presented, and the moment the mask was presented for the maximum exposure time of 153 ms. Bottom figure: results time-locked to the participants’ response, indicated by the vertical line at t = 0 s. Participants were provided with a “CORRECT” or “INCORRECT” feedback message after responding. Data were included up to 0.4 s after the participant provided a response; therefore, the number of data points near the end of the top figure (t = 1.9 s) or the beginning of the bottom figure (t = –1.5 s) is reduced (n = 8163 for correct responses, n = 2705 for incorrect responses).
Figure 4
Response accuracy (i.e., percentage of responses that were correct) versus the percentage of trials with blinking at the level of participants (N = 147), per 220 ms of elapsed time into the trial. Also shown is a least-squares regression line, means and standard deviations of the percentage of trials with blinking, and the Pearson correlation coefficients (r) and Spearman rank-order correlation coefficients (ρ). The fixation marker onset occurs at t = 0 s. The onset of the Pi-figure occurs at t = 1.1 s.
Figure 5
Mean percentage of all trials in which the participant was blinking, per group of 10 trials (11044 trials in total). Top: results time-locked to the stimulus (occurring at t = 1.1 s). An inset is provided for elapsed times between 1.0 s and 1.4 s. Vertical lines are shown for the moment the fixation marker (X) disappeared, the moment the Pi-figure was presented, and the moment the mask was presented for the maximum exposure time of 153 ms. Bottom: results time-locked to the participants’ response, indicated by the vertical line at t = 0 s. Participants were provided with a “CORRECT” or “INCORRECT” feedback message after responding. Data were included up to 0.4 s after the participant provided a response.
Figure 6
Follow-up experiment: Experience curves as a function of trial number, where Trial 1 is the first IT stimulus presented, and Trial 80 is the last IT stimulus presented. Left = Response accuracy (i.e., percentage of 159 participants who provided a correct response). Right = Median response time among 159 participants, together with 10th and 90th percentiles.
Figure 7
Follow-up experiment: Response accuracy and median response time as a function of exposure time of the Pi-figure. The median response time was calculated per participant per exposure time and subsequently averaged over the 159 participants. The means and standard deviations are provided in Table S2.
Table 2
Follow-up experiment: Inspection time task performance per self-reported illusion (N = 159).
| Response accuracy (% of trials) | Median responsetime (ms) | |
|---|---|---|
| Mean (SD) | Mean (SD) | |
| Moving/stretching illusion (n = 49) | 85.65 (7.76) | 629 (176) |
| Flash illusion (n = 39) | 84.70 (8.50) | 609 (217) |
| Black gap illusion (n = 8) | 87.81 (4.47) | 549 (81) |
| Other illusion (n = 25) | 87.04 (9.40) | 530 (104) |
| No illusion (n = 38) | 86.90 (6.15) | 552 (129) |
| Welch’s test | Welch’s test | |
| Moving/stretching vs. no illusion | t(85.0) = 0.83, p = 0.406 | t(84.7) = 2.33, p = 0.022 |
| Flash vs. no illusion | t(69.3) = 1.30, p = 0.196 | t(62.2) = 1.40, p = 0.167 |
| Black gap vs. no illusion | t(13.3) = 0.49, p = 0.632 | t(15.5) = 0.09, p = 0.932 |
| Other vs. no illusion | t(37.5) = 0.07, p = 0.946 | t(58.3) = 0.75, p = 0.457 |
Figure 8
Follow-up experiment: Percentage of trials in which participants were blinking, for each time sample. A distinction is made between trials where participants provided a correct response (n = 10915) and trials where participants provided an incorrect response (n = 1768). Top figure: results time-locked to the stimulus (occurring at t = 1.0 s). Vertical lines are shown for the moment the fixation marker (X) disappeared, the moment the Pi-figure was presented, and the moment the mask was presented for the maximum exposure time of 153 ms. Bottom figure: results time-locked to the participants’ response, indicated by the vertical line at t = 0 s. Participants were provided with a “CORRECT” or “INCORRECT” feedback message after responding. Data were included up to 0.4 s after the participant provided a response; therefore, the number of data points near the end of the top figure (t = 1.9 s) or the beginning of the bottom figure (t = –1.5 s) is reduced (n = 7967 for correct responses, n = 1569 for incorrect responses).
Figure 9
Follow-up experiment: Response accuracy (i.e., percentage of trials with a correct response) versus the percentage of trials with blinking at the level of participants (N = 159), per 220 ms of elapsed time into the trial. Also shown is a least-squares regression line, means and standard deviations of the percentage of trials with blinking, and the Pearson correlation coefficients (r) and Spearman rank-order correlation coefficients (ρ). The fixation marker onset occurs at t = 0 s. The onset of the Pi-figure occurs at t = 1.0 s.