The data in this paper pertain to those analyzed as part of a pilot monitoring study as described by Courtney et al. [1]. While the study reported on the results for three languages, the paper at hand references the results for just one study involving the Kazakh medium-of-instruction. The first variable in that data includes is ‘Student_id’, which provides a unique anonymized student ID number for each student. The next six variables in the data vary between-schools: (1) ‘experiment’, for which each school is defined as either a pilot (experiment) or control school; ‘Region’, defining the region that each school is situated in the country; (3) School_name, the name assigned to each unique school; (4) ‘Language’ defining the medium-of-instruction used in each school; ‘School_type’ where schools are defined as either (i) small, (ii) gymnasium, (iii) lyceum, or (iv) common (common used as baseline in modelling).Thereafter, the within-school variable, ‘Gender’ is given for which each child is defined as either male or female.In addition, columns 9-92 pertain to the item-response data (for all elements, 0 = incorrect, 1 = correct) for each child on each item in the five tests.2015 Test: Items OG1 to OG5.4 (columns 9-19; link items to 2016 are OG2.2A, OG2.3A, OG23A)2016 Test: Items OG2.2 to OG11.2 (20-35; link items to 2017 are )2017 Test: Items OG11.1 to K3.6 (36-47)2018 Test: Items K4.1 to K3.3 (48-68)2019 Test: Items X19K1 to K6 (69-92)The data is filtered data. Only students who completed all five literacy assessments (2015 to 2019) were included in the data. Alternative approaches for dealing with missing data (attrition) in the analysis described in this article. Original data in Cyrillic used to showcase the R software’s capacity to manage and recode such data.

THIS DATASET IS ARCHIVED AT DANS/EASY, BUT NOT ACCESSIBLE HERE. TO VIEW A LIST OF FILES AND ACCESS THE FILES IN THIS DATASET CLICK ON THE DOI-LINK ABOVE