Comprehension

Scarborough (2001) proposed that skilled reading is like a rope made from two intertwined strands. The first strand is word recognition, which comprises phonological awareness, decoding (phonics), and sight recognition of high-frequency words. The second strand is language comprehension, which includes background knowledge, vocabulary, language structures, verbal reasoning, and literacy knowledge. Both strands must be strong for the rope to hold. A learner who can decode accurately but has limited vocabulary and background knowledge will read words without understanding them. A learner with rich language comprehension but weak decoding will be unable to access the text at all. Assessment must identify which strand is the weak point before you can intervene effectively.

Reciprocal teaching, developed by Palincsar and Brown (1984), is a structured dialogue approach in which learners take turns leading a discussion about a shared text using four comprehension strategies: predicting (what will happen next?), questioning (what questions does this text raise?), clarifying (what words or ideas are unclear?), and summarising (what are the key points so far?). The teacher initially models all four roles, then gradually hands responsibility to learners. This progressive release of responsibility is what gives the approach its power: learners internalise the strategies by using them, not just hearing about them. Research consistently shows effect sizes of 0.74 to 0.88 for reciprocal teaching (Rosenshine and Meister, 1994), making it one of the most robust comprehension interventions available.

Inference is the process of using what is in the text plus what you already know to reach a conclusion the author has not stated directly. Teaching it explicitly involves three steps. First, model the process: read a passage aloud, think aloud about the gap between what the text says and what it implies, then name the inference you have made. Second, provide a frame: "The text says... I already know... so I can infer..." gives learners a scaffold to structure their thinking. Third, build the knowledge base: learners cannot infer from a text about volcanoes if they have no knowledge of how volcanoes work. Vocabulary and background knowledge instruction therefore supports inference capacity directly. Nation and Snowling (1998) showed that poor comprehenders are distinguished from good comprehenders most clearly by their ability to make inferences, not by decoding.

Guided reading involves a teacher working with a small group of learners who are reading at a similar level. The teacher introduces the text, listens to learners read, and uses targeted prompts to support both decoding and comprehension. It is most common in KS1 and lower KS2. The evidence for guided reading is mixed. When it is well structured, with the teacher actively teaching strategies rather than simply listening to reading, it can be highly effective. When it becomes passive listening, the impact diminishes. The EEF (2021) found that small group reading tuition shows an effect size of around 0.4. The key is that the teacher must be the active ingredient: prompting, questioning, and connecting the text to learners' existing knowledge rather than simply correcting errors.

Vocabulary is one of the strongest predictors of reading comprehension. Nagy and Scott (2000) estimated that learners need to know between 90 and 95 per cent of the words in a text to comprehend it adequately. If a learner encounters too many unknown words, the cognitive load of guessing at meanings overwhelms comprehension of the whole passage. Beck, McKeown and Kucan (2002) proposed a useful framework for vocabulary instruction based on three tiers: Tier 1 (everyday words), Tier 2 (academic words that appear across subjects, such as "consequence" or "significant"), and Tier 3 (subject-specific technical terms). Pre-teaching Tier 2 and Tier 3 vocabulary before reading a text is one of the most straightforward and effective ways to support comprehension, particularly for disadvantaged learners and EAL learners who may lack exposure to academic vocabulary outside school.

Disciplinary literacy is the idea that reading in history, science, mathematics, and English each requires different skills. A historian reads a primary source by questioning its provenance, bias, and context. A scientist reads a methods section to evaluate experimental design. A mathematician reads a problem to extract the underlying logical structure. Shanahan and Shanahan (2008) argued that general reading skills are necessary but not sufficient for academic success: learners also need to learn the specific reading practices of each discipline. This means that every teacher is, in some sense, a reading teacher. History teachers who model how historians interrogate sources, and science teachers who walk through how to read a data table, are building disciplinary literacy alongside subject knowledge.

EAL learners face a double challenge: they are acquiring English at the same time as being expected to use it to learn. Krashen (1982) proposed that language acquisition happens when learners encounter input that is slightly beyond their current level, what he called i+1. For reading comprehension, this means carefully matching texts to current language level while providing support that makes slightly more difficult texts accessible. Pre-teaching key vocabulary is the single most impactful strategy. Graphic organisers that make text structure visible reduce the cognitive load of navigating an unfamiliar genre. Bilingual glossaries and first-language previewing (reading about the topic in the home language before engaging with the English text) also support comprehension significantly. Paired reading with a more fluent peer or adult remains one of the most consistently effective approaches for EAL learners at all stages.