Skip to main content
. 2018 Nov 5;8:16310. doi: 10.1038/s41598-018-31435-0

Figure 5.

Figure 5

Performance on MNREAD Versus RPG Sentences. The reading performance of subjects on the embedded MNREAD sentences versus new Rassia-Pezaris-Gutenberg (RPG) sentences (i.e., the remainder) was not significantly different for either (A) accuracy (p = 0.9) or (B) speed (p = 0.5) on a Wilcoxon rank sum test across the population. Data (dark blue points) are shown for 1-cycle means (average performance over the 20 conditions, presented in randomly interleaved balanced cycles).The overall means (light blue circles) of each group (MNREAD vs RPG: accuracy, 90.8 ± 7.1% vs 91.2 ± 6.4%; speed, 69.8 ± 8.5 WPM vs 71.2 ± 7.5 WPM) do not deviate significantly from the identity line (light blue bars of standard deviation orthogonal to the line of identity). Fitted regression lines (purple) do suggest there is a trend for MNREAD sentences to have a broader amount of variability than the new RPG sentences: periods where subjects were reading faster overall read MNREAD sentences more quickly than the new RPG sentences, and periods where subjects were reading more slowly overall had more difficulty with MNREAD versus new RPG sentences, but the result is not significant (the identity line is within the 95% confidence range of the fitted regression).