The study was initiated to implement and evaluate three models of support to determine what model/s could improve the acquisition of literacy skills among learners and at what cost. Specifically, the study was designed to test interventions that could be implemented at scale.

Name of Organisation/service provider: Department of Basic Education, Office of the Director-General
Duration of Project: 2015 – December 2017
Province: North West Province

1.1 Introduction

In 2015, the Department of Basic Education (DBE) implemented and evaluated the Early Grade Reading Study to test three different intervention models, in the North West Province, with the potential to improve leaner performance in early grade literacy, to inform its national strategy and campaign on reading and literacy. The three models under evaluation are:

i) A structured teacher support programme with centralised training,
ii) A structured teacher support programme with specialist on-site coaching,
iii) A parental intervention programme.

The Early Grade Reading Study provides training and support to ensure that teachers are present, skilled and motivated to deliver systemic instruction in Setswana literacy; provides access to suitable graded materials, and expanded opportunities for learners to read. The focus of the project is Foundation Phase classrooms. When the project was launched in 2015 the focus was on the provision of support at Grade 1 level and in 2016, support was provided in Grade 2. In 2017, the project will be concluded with the provision of Grade 3 support.

The DBE is implementing and evaluating the project with support from a consortium of funders. The funder group comprises the Zenex Foundation, North West Department of Education, World Banks 3IE, UNICEF and Anglo American. The focus of the Zenex Foundation’s support is on the two school-based models focusing on teacher development.


1.2 Theory of Change

The theory of change suggests that the acquisition of reading comprehension in the Foundation Phase is the gateway to all learning and is dependent on the mastery of vocabulary and decoding. It maintains that whilst vocabulary can come naturally through speaking and hearing, decoding does not come naturally and should be taught systematically. Furthermore, the acquisition of decoding skills requires reforming the instructional core, which refers to the relationship between the teacher, learner and resources as drivers of educational change.

Below is a description of the two intervention models supported by the Zenex Foundation:

Intervention 1: Coaching and Resources

The testable hypothesis is that improved teaching and learning can be achieved through a combination of the provision of structured resources and on-site mentoring and coaching for the teachers. The key characteristics of this model are as follows:

i. Teaching and learning resources:
    a. Highly structured lesson plans,
    b. Curriculum planning and tracking tools,
    c. Graded readers in Setswana.
ii. Teacher training and coaching:
    a. Orientation: 2-day dosage and 4, half –day orientation per term.
    b. On site coaching: 10-day dosage per teacher per year.

Intervention 2: Centralised Training and Resources

The testable hypothesis is that the provision of high quality structured resources and centralised training on the use of teaching and learning resources will lead to improved teaching and learning. This model excludes coaching. The key characteristics are as follows:

i. Teaching and learning resources:
   a. Highly structured lesson plans,
   b. Curriculum planning and tracking tools,
   c. Graded readers in Setswana.
ii. Teacher Training:
    a. 8-day dosage per year


1.3 Methodology

A Randomised Control Trial design is used to evaluate the impact of the two interventions. The sample consists of 100 schools (50 schools for each model) in quintile 1 – 3 and an additional 80 control schools in the same quintile in two Districts in the North West Province.

The study makes a comparison between schools receiving the two intervention models and the control schools. It further makes a cost-benefit analysis to establish the most cost-effective intervention.

Researchers used questionnaires, interviews, observations, artifacts and document analysis to collect data on schools characteristics and teachers. The Early Grade Reading Assessment instrument is used collect data on learners’ literacy competences.


1.4 Preliminary Findings from the Midline Report

The midline report was compiled using learner test data collected at the end of 2015 and the data on teachers was collected from 2015 to 2016. The baseline report gives insight on the impact of both interventions on learners after a year of implementation in Grade 1 and teacher practices over two years of implementation.  The key findings from this report are as follows:

1.4.1 Impact of the two models on learner performance:

• Nine months into the programme during Grade 1, the two intervention models showed a slight impact on reading proficiency when compared to control schools.
• The best-performing and worst-performing learners did not benefit as much as learners in the upper-middle range from both intervention     models.
• The coaching model yielded greater learner gains compared to the centralised training model. However, the centralised training model was found to be more cost effective when both cost and impact were taken into account.

1.4.2 Impact of intervention models on teacher practices

• There was evidence of pervasive lesson planning and improved curriculum coverage from both models.
• There was better classroom management from teachers in both intervention models, which was evident in the way teachers streamed learners according to their reading proficiency.
• There was evidence of more frequent individualised reading assessments of learners from both models compared to control schools.
• The intervention was most effective with teachers who performed better on the baseline comprehension test. This implies that teachers may need a certain level of competence to apply the scripted lesson plans effectively.


1.5 Conclusion

The study is important in informing literacy support at for various reasons, including:

• It is important to understand in more detail what works and what does not, particularly as it pertains to working at scale.
• There is a need to better understand how the use of the structured learning approaches (lesson plans, readers for learners and training/coaching) affects classroom practices and impacts on learning outcomes.
• The study can provide important insights into the trajectory of learners as they move from Grade 1 to Grade 3 and contribute to a better understanding of how learners make the shift from ‘learning to read to reading to learn’.


Read 22 times