Fig. 1: Performance on four reading subskill measures for the first three instructional years. | Nature Human Behaviour

Fig. 1: Performance on four reading subskill measures for the first three instructional years.

From: Inadequate foundational decoding skills constrain global literacy goals for pupils in low- and middle-income countries

Fig. 1: Performance on four reading subskill measures for the first three instructional years.

a–d, The data from the four subskill measures are shown: letter name identification (a), letter sound identification (b), non-word reading (c) and oral reading fluency (d). For each plot, the number of subskill scores, countries and languages analysed is reported under the name of the task. Each plot shows the jittered raw data (the dots representing average subskill scores); the four quartiles of the ordered data (the box-and-whiskers plots) with the grey horizontal lines representing the medians (50%), the bounds of the boxes representing the lower and upper 25% quartiles, and the whiskers representing the expected variation of the data; and the estimated data distribution (the clouds) together with the means (the dots) and the 95% confidence intervals (the error bars). The DIBELS benchmarks (where available) for substantial and severe risk are represented using dashed lines (black and red, respectively) averaged across three time points (beginning, middle and end of year). The number of data points in each plot varies because not all EGRA surveys included all four decoding measures.

Back to article page