Fig 1.
Description of the Courier (top-row), Eido (middle-row) and DejaVu (bottom-row) fonts used in our experiments.
Note that the letter x-height and center-to-center spacing were the same for the different fonts.
Fig 2.
Description of the experimental protocol for the gaze-contingent reading experiment (Exp1 and 5, plot a), letter recognition experiment (Exp 2 and 6, plot b) and word recognition experiment (Exp 3 and 4, plot c). In Experiments 1 and 5 (plot a), the subject fixated a dot, pressed a button to display the text (and the gaze-contingent central scotoma centered on the foveal fixation), then made several fixations to identify the words of the sentence while they were reading them aloud (The black circle around the scotoma displayed on the figure was not visible during the experiments and the scotoma was a gray mask with the same luminance as the background). The subject pressed the button when he/she identified the presented words. In Experiments 2, 3, 4 and 6, the subject fixated a dot, pressed a button to display the letter trigram (Experiments 2 or 6, plot b) or the word/pseudoword (Experiments 3 or 4, plot c) in the lower visual field. In Experiments 2, 3 and 6, the stimulus disappeared after the preset presentation time and was replaced by a backward mask (200 ms). In Experiment 4, the stimulus disappeared after the subject pressed one of the two possible buttons.
Fig 3.
Description of the typical improvement of reading speed in function of letter print-size.
First, data are fitted with an exponential function. The maximum reading speed (MRS) corresponds to the asymptote of the curve. The critical print size (CPS) corresponds to the size inducing a reading speed of 90%. The reading acuity (RA) is defined as the intersection between the function curve and the abscissa axis.
Fig 4.
Sentence reading with simulated scotoma (Experiment 1).
Effect of print-size (x-height, logMar) on reading speed (words per minute) for each of the six subjects. For both fonts, the '+' on the ordinate axis represents the maximum reading speeds (MRS) and the 'x' on the abscissa axis represents the Critical Print Size.
Fig 5.
Sentence reading with simulated scotoma (Experiment 1).
Effect of font type on the three extracted reading parameters: reading acuity (RA), Critical Print Size (CPS) and log Maximum Reading Speed (logMRS).
Table 1.
Fixed effects results of the linear mixed-effects model in Experiment 1 (Dependent Variable is Maximum Reading Speed).
Table 2.
Fixed effects results of the linear mixed-effects model in Experiment 1 (Dependent Variable is Critical Print Size).
Table 3.
Fixed effects results of the linear mixed-effects model in Experiment 1 (Dependent Variable is Reading Acuity).
Fig 6.
Crowded letter recognition (Experiment 2).
Average number of unrecognized letters per trigram for each subject and for both fonts (red color: Eido font, blue color: Courier font). Standard deviations were obtained by bootstrapping for the distribution of errors per trigram (1,000 simulation trials). The dark part of the bars corresponds to letter errors due to letter mislocations.
Table 4.
Fixed effects results of the linear mixed-effects model for Experiment 2 (Dependent Variable is the number of letter errors).
Table 5.
Fixed effects results of the linear mixed-effects model for Experiment 2 (Dependent Variable is the number of mislocalized letters).
Fig 7.
Word recognition (Experiment 3).
Word recognition error rate for each subject, for the Eido font (red) and the Courier font (blue).
Table 6.
Fixed effects results of the generalized linear mixed-effects model for Experiment 3 (Dependent Variable is the word recognition error rate).
Fig 8.
Letter recognition error rate (word recognition task) vs. Letter recognition error rate (letter recognition task).
Letter recognition error rate during the word recognition task (Experiment 3) as a function of letter recognition error rate during the letter recognition task (Experiment 2), averaged across subjects for each letter and both fonts. For instance, the position of the letter 'n' for the Courier font shows that the letter recognition error rate was about 0.31 for the letter recognition task and 0.27 for the word recognition task when the Courier font was used.
Table 7.
Fixed effects results of the generalized linear mixed-effects model to predict the proportion of recognition errors for each letter during the word recognition task.
Fig 9.
Word lexical decision (Experiment 4).
Average and standard deviation of the log lexical decision time for each subject and for both fonts (red color: Eido font, blue color: Courier font).
Table 8.
Fixed effects results of the linear mixed-effects model for Experiment 4 (Dependent Variable is the log lexical decision reaction time).
Fig 10.
Sentence reading with simulated scotoma (Experiment 5).
Average and standard deviation log reading speed for each subject and for both fonts (red color: Eido font, blue color: Courier font).
Table 9.
Fixed effects results of the linear mixed-effects model for Experiment 5 (Dependent Variable is the log reading speed).
Fig 11.
Letter recognition (Experiment 6).
Average number of unrecognized letters per trigram for each subject and for both fonts (red color: DejaVu font, blue color: Courier font). Standard deviations were obtained by bootstrapping for the distribution of errors per trigram (1,000 simulation trials). The dark part of the bars corresponds to letter errors due to letter mislocations.
Table 10.
Fixed effects results of the linear mixed-effects model for Experiment 6 (Dependent Variable is the number of letter errors).