S1 File - A crowding free digital interface to help French-speaking children learn to read

S1.Text. Testing materials. S2.Text. Comparison of the experimental group (Group 1 & Group 2) vs. control group (children in the same class as the experimental group but who did not participate in the training sessions). The active control group was trained with the classical school curriculum by their teacher (n = 4 classes). Teacher participation was voluntary. The 4 teachers who participated were equally experienced (mid-to-late stages of careers). All children were assessed by teachers on the French National reading evaluations, in September and January, and by the research team on reading evaluations (pseudoword reading and meaningful text reading) in June. Only one teacher agreed to share the data from his class on the national evaluation (n = 24, 8 in the control group and 16 in the experimental group).S3.Text. Digital saccade as a function of fluency level during pseudowords reading. S4.Text. Digital saccades as a function of reading fluency performance for meaningless text (Alouette). S5.Text. Finger movement as a function of reading fluency performance for meaningful text (Monsieur Petit). S1.Table. Training stimuli. Overall structure and choice of pseudowords stimuli. In each phase, 20 monosyllabic, 60 bisyllabic and 40 trisyllabic pseudowords were introduced. The complexity of each pseudowords ranged from 1) simple graphemes-simple syllables (CV structure) to 2) complex graphemes-simple syllables (CV structure) and finally 3) complex syllables (CVC or CCV structure). S2.Table. Reading results of control study. Mean and SD for each reading-related skill and each modality (Paper or tablet) were presented, along with p-value of two-sample t-test. S1.Fig. Data of finger kinematics were collected during 6 training sessions (at phase 1 for Group 1 and phase 2 for Group 2). Within each group, children were divided into good or bad decoders based on their decoding score using BELO test. S2.Fig. Change in digital saccade variables according to text decoding level (good vs. poor decoders) and session (1–6) for Group 2. A. Average speed of digital saccade (mm/s). B. Average length of digital saccades (mm). C. Proportion of digital regressive saccades. D. Average duration of each digital fixation in milliseconds. Error bars represent standard error estimated by LMM. Significance levels:. 0.1 S3.Fig. Example of syllable recognition test. Children had to circle the syllable ‘VI’ among six elements proposed underneath. S4.Fig. Procedure with a control group included. Along with 54 children of the experimental group, a control group of 19 children from the same classes but did not participate in training were tested in June by our research team. Furthermore, a subset of data from the two national evaluations (in September and January) were collected. S5.Fig. Change in finger movement according to pseudowords reading fluency level (good vs. poor readers) and session (1–6) for Group 1. A. Average speed of digital saccade (mm/s). B. Average length of digital saccades (mm). C. Proportion of digital regressive saccades. D. Average duration of each digital fixation in milliseconds. Error bars represent standard error estimated by LMM. Significance levels:.0.1 S6.Fig. Change in finger movement according to reading fluency level for meaningless text (good vs. poor readers) and session (1–6) for Group 1. A. Average speed of digital saccade (mm/s). B. Average length of digital saccades (mm). C. Proportion of digital regressive saccades. D. Average duration of each digital fixation in milliseconds. Error bars represent standard error estimated by LMM. Significance levels:.0.1 S7.Fig. Change in finger touch variables according to meaningful text reading fluency level (good vs. poor readers) and session (1–6) for Group 1. A. Average speed of digital saccade (mm/s). B. Average length of digital saccades (mm). C. Proportion of digital regressive saccades. D. Average duration of each digital fixation in milliseconds. Error bars represent standard error estimated by LMM. Significance levels:. 0.1 (ZIP)