Figure 1.
Individual psychometric functions of the 2-point discrimination task.
The figure shows individual psychometric functions for each of the 26 participants. The percentages of ‘distance felt’ answers across pin distances (depicted as dots) were fitted with a binary logistic regression and the spatial discrimination threshold was defined as the pin distance closest to the 50% crossing of the fitted sigmoid curve. Error bars are confidence limits derived from bootstrapping 1999 curve fitting simulations. Subjects 3, 4, 11, 13, 14, 15, and 21 were excluded from further analyses (see Materials and Methods).
Figure 2.
Conditional probabilities for same and different subsequent perceptual decisions.
Bar graphs show conditional probabilities for the decision ‘distance felt’ in combination with either the same (dark gray bars) or a different (light gray bars) decision on the preceding trial. The left panel shows probabilities across all five current and preceding pin distances; the right panel shows probabilities for each of the currently presented pin distances in combination with all other pin distances. Whiskers represent the standard error of the mean. Significant differences between same and different decisions are denoted with a star for p<0.05. See also Table 1.
Table 1.
Wilcoxon signed rank tests for conditional probabilities of same versus different subsequent decisions.
Figure 3.
Nonparametric regression analysis for same and different decisions for 20 time lags.
Each point shows delta, that is, the conditional probability for the decision ‘same’ minus the conditional probability for the decision ‘different’ for each participant. Probabilities for ‘same’ and ‘different’ decisions were computed for the decision on the current distance in combination with a previous decision up to a lag of 20. The solid line shows the regression line and represents the mean values of delta for the different lags as inferred from the data of the participants. The light gray shading indicates the region of two standard deviations around the regression line and includes the estimated variability across subjects. The dark gray shading indicates the variability within two standard deviations of the posterior delta values that is left when the variability across subjects is accounted for. See Materials and Methods for details.
Table 2.
Descriptive statistics and statistics for Wilcoxon signed rank tests for explained variance of pin distances and decision hysteresis.
Table 3.
Statistics for Wilcoxon signed rank tests for the comparison of explained variance of pin distances and decision hysteresis at ‘threshold’ versus the other four current pin distances.
Figure 4.
Explained variance of pin distances and decision hysteresis.
Bar graphs show the amount of explained variance of participants’ decisions for stimulus property (i.e., presented pin distance; dark gray bars) and decision hysteresis (light gray bars). The left panel shows explained variances across all five current and preceding pin distances; the right panel shows explained variances for each of the currently presented pin distances in combination with all other pin distances. Whiskers represent the standard error of the mean. Significant differences between variances explained by pin distances and hysteresis are denoted with one star for p<0.05, with two stars for p<0.01 and with three stars for p<0.001. See also Table 2.
Figure 5.
Explained variance of pin distances and decision hysteresis at threshold.
The scatterplot shows explained variance of pin distances (dark gray circles) and decision hysteresis (light gray circles). At threshold we revealed a trend showing that hysteresis explains more variance than the stimulus (i.e., the applied pin distance).
Figure 6.
Spearman correlation analysis between explained variance of pin distances and decision hysteresis.
The scatterplot shows the negative correlation between explained variance of pin distances and decision hysteresis. This suggests that the less variance was explained by the stimulus (i.e., the applied pin distance), the more influential hysteresis became for decisions.