Figure 3

(a) Individual average success rates in the forward vs. backward conditions: individual average success rates plotted against chance level (11%) in forward (pink) and backward (turquoise) conditions. Each colored dot represents the average success rate of one participant. Chance level is represented by the dashed gray line. All participants in both groups performed well above the chance level. The median of each group is represented by a solid black line. The two individual accuracy distributions closely resemble each other, further suggesting entirely comparable performance in the forward vs. backward fields. (b) Equivalence testing using the delta (\(\Delta\)) distribution (forward–backward performance). The grey columns represent the frequency of values of the delta distribution (i.e., the difference in percentage points between forward and backward performances). The dashed black lines represent the upper and lower equivalence bounds \({(\Delta }_{\mathrm{L}}=-0.024\) and \({\Delta }_{\mathrm{U}}=0.024)\). The thick blue line represents the 90% confidence interval (CI) of the \(\Delta\) distribution [− 0.005, 0.02]. The thin blue line represents the 95% CI of the \(\Delta\) distribution [− 0.01, 0.02]. The black square represents the median of the \(\Delta\) distribution (M = 0.0055). As we can observe, both the 90% and the 95% CIs include the value of 0, but do not include the equivalence boundaries. Thus, we can conclude that the difference between success rates distributions of the forward and backward conditions is not statistically significant and the distributions are statistically equivalent. (c) Heat maps for the average success rate in each of the nine cells across participants. Bottom left: blindfolded sighted participants in the forward condition (n = 14). Bottom right: blindfolded sighted participants in the backward condition (n = 14). In both panels, each cell on the 3 × 3 grid is color-coded based on the average success rate in it. Darker shades of blue represent higher success rates. Accuracy was very high for both groups of participants in all cells. This Figure was created using R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. R version 3.5.2 (2018-12-20). https://www.R-project.org/.