South African learners consistently struggle with reading comprehension, performing poorly in both international and local assessments. A significant issue is that 81% of grade 4 learners (aged 9 or 10) are unable to read for meaning: they can decode words, but do not necessarily understand them.
While this problem has received considerable attention, no clear explanation has emerged.
In my recent PhD thesis, I considered a crucial, but often overlooked, piece of the puzzle – the curriculum policy. My research sought to uncover and understand the gaps and contradictions in reading comprehension, especially between policy and practice, in a grade 4 classroom.
This research revealed a difference between curriculum policy and practice, and between what learners seemed to have understood and what they actually understood in a routine reading comprehension task.
My main findings were that:
grade 4 learners were being asked overly simple, literal questions about what they were reading, despite the text being more complex than expected
the kinds of questions that learners should be asked (as indicated in the curriculum policy) were different from what they were being asked
this gap led to learners seeming to be more successful at reading comprehension than they actually were.
Pinpointing the gaps between what the policy says and how reading comprehension is actually taught at this crucial stage of development (grade 4) could pave the way for more effective interventions.
Curriculum policy
South African teachers are expected to base their reading comprehension instruction and assessment on the guidelines provided by the 2012 Curriculum and Assessment Policy Statement.
The policy outlines specific cognitive skill levels – essentially, ways of thinking and understanding – that learners should master for each reading task. These levels are drawn from Barrett’s 1956 Taxonomy of Reading Comprehension, an international guideline. It’s based on the popular Bloom’s Taxonomy of Reading Comprehension, which categorises reading comprehension according to varying skill levels.
According to Barrett’s Taxonomy, reading comprehension involves five progressively complex levels:
Literal comprehension: Identifying meaning that is directly stated in the text. (For example, “Name the animals in the story”.)
Reorganisation: Organising, paraphrasing, or classifying information that is explicitly stated. (“Find four verbs in the story to describe what the animals did.”)
Inference: Understanding meaning that is not directly stated, but implied. (“When in the story is the leopard being selfish?”)
Evaluation: Making judgements about the text’s content or quality. (“Who do you think this story is usually told to?”)
Appreciation: Making emotional or personal evaluations about the text. (“How well was the author able to get the message across?”)
Typically, reading comprehension tasks will assess a range of these cognitive skills.
South Africa’s Curriculum and Assessment Policy Statement document specifies (on pages 91-92) that all reading comprehension tasks should comprise questions that are:
40% literal/reorganisation (lower-order thinking skills)
40% inferential (middle-order)
20% evaluation and appreciation (higher-order).
This approach aims to allow most students to demonstrate a basic understanding of the text, while challenging more advanced learners.
However, as my classroom case study shows, the system appears to be failing. There was a mismatch between the policy and what was taking place in the classroom.
Classroom practice
For this research, I observed the reading comprehension practices in a single classroom in a public school in the Eastern Cape province. This took place over six months, at a time when schools were not fully reopened during the COVID-19 pandemic.
The task in question included a text and activity selected by the teacher from a textbook aligned with the policy. My analysis (which used Appraisal, a linguistic framework that tracks evaluative meaning) showed that most of the text’s meaning was implicit. To fully understand it, learners would need higher-order thinking and sophisticated English first-language skills. This was a surprising finding for a grade 4 resource, especially because most learners in this study were not English first-language speakers.
Even more surprising, learners achieved seemingly high marks on comprehension, with an average of 82.9%. This suggested they understood this complex text.
However, I found that the questions in the textbook did not align with policy. Instead of the balance of skills required by the policy, 73% of the questions called only for lower-order skills. Only 20% were inferential and a mere 7% required evaluation or appreciation (middle- to higher-order skills).
At least six of the 15 available marks could be gained simply by listing explicitly stated items, not requiring genuine comprehension.
This reveals that, in this classroom, activities labelled as policy-compliant actually tested only lower-order comprehension. Learners could pass simply by identifying and listing information from the text. This creates a false sense of comprehension success, as revealed by the high marks.
When learners were tested on the same text but using different questions that I designed to align with the policy requirements, they scored lower marks, especially for the higher-order questions.
This mismatch might partly explain why South Africans score poorly in international tests (which require more higher-order thinking).
Why this matters and moving forward
These findings are concerning, as learners may be lulled into believing that they are successful readers. A false sense of accomplishment could have significant impacts on the rest of their education.
Comprehension difficulties can’t be blamed solely on the disconnect between policy and practice, however. Many other contextual factors shape how learners perform in reading comprehension tasks.
In my study, factors like COVID-19, insufficient home language teaching policies, educational inequalities, and the pressures on teachers during a crisis (brought on by COVID-19) all contributed to the literacy crisis.
Read more: South Africa’s reading crisis: focus on the root cause, not the peripherals
However, two key points became clear during this study.
Firstly, teaching materials favour lower-order comprehension skills, skewing perceptions of learners’ abilities.
Secondly, teachers may lack the knowledge, resources or motivation to adjust these materials to truly align with the national policy in how reading comprehension is assessed.
This calls for urgent intervention in how reading comprehension is taught and assessed and in how teachers are prepared to do this effectively.
