Skip to main content

Early-grade reading support in rural South Africa: A language-centred technology approach


This article describes a short-term longitudinal study conducted in low-performing rural primary schools in South Africa. The authors were involved in developing a two-year multimedia reading programme for rural South African children in grades 1–3, and sought to assess key learning outcomes in existing school computer laboratories. The programme provided interactive local-language support in three local languages plus English, with content designed to be culturally appropriate and contextually relevant for the target group. The study tracked a total of 215 learners, some of whom used the programme once a week for its duration, while a control group did not. Findings revealed a positive and significant impact on local-language reading fluency and reading comprehension among the children who had used the instructional software. The outcome demonstrates nearly three-quarters of a year of additional reading growth compared to reading growth at control schools within the sample, as well as a twofold increase in reading comprehension in the treatment group. The study demonstrates that technology-based reading support can help rural learners in South Africa to make substantive gains so they can complete their schooling. The ability to accomplish a full cycle of primary school with well-developed reading skills has important implications for lifelong learning. The authors argue that guided and contextualised instructional software can lead to enhanced learning outcomes compared to unguided digital content or traditional teacher support alone.


Soutien précoce à la lecture en Afrique du Sud rurale : approche technologique et langagière – Cet article décrit une étude longitudinale menée à court terme dans des écoles primaires rurales peu performantes d’Afrique du Sud. Impliqués dans la conception d’un programme biennal multimédia de lecture destiné aux élèves des trois premiers niveaux primaires habitant l’Afrique du Sud rurale, les auteurs ont tenté d’évaluer les principaux résultats d’apprentissage dans les salles d’informatique propres aux écoles. Ce programme apportait un soutien interactif dans trois langues locales et en anglais, le contenu ayant été conçu de sorte à être adapté à la culture et pertinent pour le groupe cible. L’étude a suivi un total de 215 écoliers, dont une partie a utilisé le programme une fois par semaine pendant sa durée, un groupe témoin n’y avait pas accès. Les conclusions établissent chez les élèves exposés au didacticiel un impact positif et significatif sur la fluidité de la lecture en langue locale ainsi que sur la compréhension écrite. Le résultat démontre une progression supplémentaire en lecture de presque trois trimestres, comparée aux progrès des écoles témoins dans l’échantillon, ainsi qu'un doublement en compréhension écrite pour le groupe traité. L’étude prouve qu’un soutien technologique à la lecture peut aider les élèves en Afrique du Sud rurale à en tirer d’importants avantages jusqu’à pouvoir achever leur scolarité. La capacité à accomplir le cycle entier de l’enseignement primaire avec de bonnes compétences en lecture a des conséquences décisives pour l’apprentissage tout au long de la vie. Les auteurs affirment qu’un logiciel pédagogique guidé et contextualisé peut améliorer les résultats d’apprentissage, comparé à un contenu numérique non guidé ou au seul soutien traditionnel de l’enseignant.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others


  1. The South African Department of Basic Education prefers the term “home” language to the internationally used terms local language or mother tongue (DBE 2013).

  2. “The South African national system of Adult Basic Education and Training (ABET) was established shortly after the first South African Democratic elections in 1994, with the introduction of the ‘National adult basic education and training framework: Interim guidelines’ in September 1995. Subsequently, the Adult Basic Education and Training Act was promulgated in 2000” (McKay 2015, p. 366). For more information on ABET, see [accessed 12 April 2019].

  3. According to the initiative’s website, All Children Reading: A Grand Challenge for Development was “launched in 2011 by the US Agency for International Development (USAID), World Vision and the Australian Government, [It] is a series of [grant] competitions that leverages science and technology to source, test, and disseminate scalable solutions to improve literacy skills of early grade learners in developing countries” ( [accessed 28 March 2019]).

  4. Foundation-phase learners are students in grades 1–3.

  5. The software was developed with all languages loaded onto one programme instead of separate versions and installed at each treatment site in all four languages, regardless of the local home language. An important component of the software was to provide the material in multiple languages so that learners could toggle between the language of their choice, at their own pace. Given the regional multilingual context, we had anticipated this to play a larger role in the research, but found in reality that learners mainly stuck to their respective home languages.

  6. We prepared 10 lessons for grade 2, and 15 lessons for grades 3 and 4 respectively.

  7. For the English version, see RSA (2011d).

  8. There were no significant differences in grade advancement between treatment and control schools.

  9. These measure the difference in EGRA component score from the learner’s previous score.

  10. We initially anticipated a randomised controlled trial with stratified random sampling at the school level. However, the randomisation revealed heterogeneity between treatment groups at baseline. Therefore, we applied the indicated adjustments in our analysis to correct for baseline differences.

  11. See Castillo (2017) for further details regarding statistical models and instrument validation.

  12. In addition to improving efficiency in estimating change over time, fixed effects methods offer the added benefit of controlling for all unchanging characteristics of the individuals, whether observed or unobserved (Allison 2009). Within a fixed effects model, variance associated with school clustering and time-invariant background characteristics are accounted for when estimating the treatment effect.

  13. “This method first subtracts individual scores of the time point of interest from the same individual’s score at the T-1 timepoint. The overall mean of the treatment group difference score is then subtracted from the overall mean of the comparison group difference score to assess BFI impact. This double differencing method is a widely-used technique for accounting for observed heterogeneity at baseline” (Castillo 2017, pp. 39–40, referring to Allison 2009).

  14. For discussion of year 3 results, see Castillo (2017).


Download references


We thank our partners from the Molteno Institute for Language and Literacy as well as the University of Pennsylvania for their dedicated collaboration at each stage of this study. The research presented here was supported by an All Children Reading Grant from the United States Agency for International Development (USAID), World Vision and the Australian Government.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Nathan M. Castillo.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Castillo, N.M., Wagner, D.A. Early-grade reading support in rural South Africa: A language-centred technology approach. Int Rev Educ 65, 389–408 (2019).

Download citation

  • Published:

  • Issue Date:

  • DOI: