Figure 1: Four conditions fitted with individual thresholds, common (but fitted) slope and lapse rate (guess
rate fixed at 0.5). This is a 6 parameter (4 thresholds, 1 slope, 1 lapse rate) model.

Figure 2: Model comparison. Model shown in Figure 1 (4 thresholds, 1 slope, 1 lapse rate) versus model
with shared threshold (1 threshold, 1 slope, 1 lapse rate). I.e., are thresholds differences 'statistically reliable'?

Figure 4: Goodness-of-Fit: Quadratic trend on threshold, common slope, common lapse rate, logistic form versus
saturated model.