Prediction turns reading into an active cycle of anticipation and checking. A reader uses titles, openings, text structure, genre, prior knowledge and new evidence to form a reasonable expectation about what may come next, then revises that expectation as the text develops.
This SETC guide covers prediction before, during and after reading; evidence strength; narrative and nonfiction prediction; genre; text structure; revision; misconceptions; examination transfer and teaching routines.
The governing rule is prediction plus verification. Being willing to change a prediction is evidence of reading, not evidence of failure.
What is prediction?
Prediction is a forward-looking inference about what a text may contain, explain or do next. It is provisional and evidence-based.
prediction is best understood as a controlled reading process rather than a decorative classroom activity. The reader notices a cue, makes a provisional decision, checks that decision against the text and changes course when evidence demands it. The process may become fast with experience, but the underlying logic remains inspectable.
Why prediction matters
Prediction creates purpose and makes mismatch visible. Readers notice whether new information confirms or changes the current mental model.
The larger educational value is transfer. A learner who can use prediction only when a worksheet names the skill has learned a task format, not a reading strategy. SETC therefore treats independent selection, evidence use, recovery and verification as part of mastery.
What prediction is not
- random guessing
- deciding the ending before reading
- treating genre expectations as proof
- refusing to revise
- predicting every tiny detail
- using prediction when a simpler strategy would work
These distinctions matter because many weak habits resemble the target skill from the outside. A page can contain marks, a reader can make a prediction, or a learner can reread several times without actually improving comprehension. The behaviour counts only when it changes the quality of the reader’s mental model.
The operating cycle
- Notice the cue or problem that makes the strategy relevant.
- State the current interpretation, uncertainty or purpose.
- Choose the smallest action likely to improve understanding.
- Check the result against nearby and wider evidence.
- Keep, revise or abandon the action based on what the text now supports.
With practice, the five steps compress into a rapid loop. Early teaching should keep them visible because visibility makes diagnosis possible. Later practice should remove prompts so the learner can run the loop without being told which step comes next.
Previewing titles and headings
Titles and headings narrow topic, scope and likely organisation before the detailed reading begins.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner predicts the kind of information a heading is likely to introduce. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix previewing titles and headings with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Using first lines
Opening sentences can establish setting, claim, problem or direction and provide an early evidence base.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner makes one broad prediction from the opening and records the clue. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix using first lines with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Predicting from genre
Genre creates probabilities about what a text is likely to do, but those probabilities remain subordinate to actual evidence.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner separates what the genre makes likely from what the passage specifically supports. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix predicting from genre with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Predicting narrative events
Goals, conflicts, obstacles and foreshadowing constrain likely next events.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner predicts one next move from a character’s current goal and obstacle. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix predicting narrative events with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Predicting nonfiction structure
Definitions, problems, causes and comparisons often forecast the kind of section that may follow.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner predicts whether the next paragraph is likely to exemplify, explain, contrast or solve. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix predicting nonfiction structure with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Using prior knowledge
Relevant background knowledge can generate expectations, but it must remain revisable.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner writes an expectation from prior knowledge and then checks what the text actually says. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix using prior knowledge with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Revising predictions
A changed prediction is a successful update when new evidence makes the old one less plausible.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner keeps the first prediction visible and writes what evidence forced revision. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix revising predictions with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Tracking confirming evidence
Prediction is useful when readers notice what supports it rather than merely waiting for the outcome.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner marks one confirming clue and one clue that weakens the expectation. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix tracking confirming evidence with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Managing competing predictions
Some texts deliberately support several outcomes at once. Mature readers can hold alternatives.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner keeps two plausible outcomes and updates their relative strength. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix managing competing predictions with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Predicting arguments
Readers can anticipate the evidence, counterclaim or consequence that a claim would logically require.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner pauses after a claim and lists what support would make it convincing. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix predicting arguments with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Micro-prediction in sentences
Grammar and collocation allow readers to anticipate what type of phrase can fit, supporting fluent parsing.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner predicts the grammatical role of a hidden sentence ending before revealing it. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix micro-prediction in sentences with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Knowing when not to predict
Expert strategy use includes abandoning prediction when direct evidence extraction, inference or close reading is more efficient.
A useful diagnostic is to ask the learner to explain exactly what changed in understanding at this point. If the learner can perform the surface action but cannot say why it helps, the strategy may be procedural without being conceptual. If the learner can explain the idea but cannot execute it on a fresh passage, practice needs to move from explanation to retrieval.
For practice, begin with a short passage where the learner compares two passages and decides where prediction earns its time cost. Require the learner to justify the decision using words, relationships or structure from the text. Then vary the topic and genre while keeping the same reading demand. This prevents the learner from tying the skill to one familiar worksheet format.
Transfer becomes visible when the learner notices the same need without a label. At that point, mix knowing when not to predict with vocabulary, inference, main-idea and evidence tasks so that strategy selection becomes part of comprehension rather than a separate exercise.
Common failure patterns
- guessing from preference
- overconfidence from little evidence
- failing to revise
- predicting too vaguely
- predicting too specifically
- ignoring structural clues
- treating prediction as a story-only skill
- spending more time predicting than reading
A wrong answer does not reveal which of these failures occurred. Teachers should therefore examine the first faulty decision, not only the final response. The same visible error can come from different causes and requires different repair.
Before reading
Use prediction to activate relevant knowledge and create a purpose without pretending to know the text already.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
During reading
Update expectations at meaningful stopping points rather than after every sentence.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
After reading
Compare early expectations with the completed text to identify which clues were genuinely useful.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
Narrative
Track goals, obstacles, causal chains and foreshadowing.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
Nonfiction
Predict organisation, evidence types and explanatory moves.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
Argument
Anticipate support, qualification and counterargument.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
Digital reading
Use headlines and section labels cautiously because click-driven titles can overpromise.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
Exams
Keep preview prediction brief and tied to answer-relevant structure.
In this context, the strategy should stay proportional to the reading purpose. More visible work is not automatically better work. The learner should use enough structure to preserve meaning, then return attention to the text rather than letting the support tool become the main task.
A useful transfer exercise is to compare two texts that require different versions of prediction. Ask the learner what stayed the same in the thinking process and what changed because of genre, difficulty, purpose or time pressure.
Worked example: Maren
Maren predicts an ending from the title and defends it after the protagonist’s goal changes. Her tutor preserves the original prediction and asks what evidence now makes a different outcome more plausible. Maren learns that updating is the purpose of the exercise.
The tutor asks Maren to verbalise the decision in a short think-aloud. This reveals which cue she noticed, what she assumed and why her first approach did or did not work. The correction then targets the first weak link instead of simply replacing her answer.
Worked example: Iona
Iona thinks prediction means inventing. She begins with nonfiction headings and learns to predict information type rather than content detail: a section called Causes of Coastal Erosion will likely explain mechanisms, not provide a personal narrative.
Iona’s practice keeps evidence close to the decision. She learns to distinguish what the text states, what she adds and what remains uncertain. That separation makes her reading more precise and makes later explanation much easier.
Worked example: Leonie
Leonie spends too long previewing exam passages. She adopts a ten-second routine: title, headings, likely structure, start reading. She revises only when a mismatch would improve comprehension.
Leonie’s version adds a time constraint. She keeps the reasoning but reduces the visible steps, using a brief notation or mental routine rather than a full written scaffold. The aim is not hurried reading; it is efficient control.
A six-stage practice architecture
Stage 1: model the strategy with a think-aloud. Stage 2: guide the learner through the decision with prompts. Stage 3: require the learner to explain the process before acting. Stage 4: remove prompts and use fresh short texts. Stage 5: interleave the strategy with neighbouring comprehension skills. Stage 6: test delayed transfer under realistic time conditions.
Do not advance only because a learner completes one set correctly. Repeat after a delay and change surface features. If accuracy disappears, the earlier success may have depended on memory for the format. If accuracy remains but speed collapses, retain the same logic and build fluency through shorter timed retrieval drills.
Error logs and correction loops
Record the decision that caused the error rather than the question number alone. Examples include ignored heading, unsupported assumption, failure to update, excessive marking, vague reference, or rereading without a target. After several sessions, recurring categories reveal what deserves focused repair.
A correction loop should include a fresh retest. Seeing the model answer is not evidence of learning. The learner should perform the repaired process on a new passage where the wording and topic differ.
Retrieval and spacing
Short delayed practice is more informative than immediate repetition. Ask the learner to recall when the strategy is useful, what cue triggers it and how to verify the result before opening notes. This tests whether the operating rule can be reconstructed from memory.
Spacing also reduces the illusion of mastery created by familiarity. A strategy that survives a gap is more likely to be available during independent reading and examinations.
Cross-curricular transfer
prediction can support reading in Science, Humanities, Mathematics word problems and digital information environments. The content changes, but the control problem remains: the reader must build a coherent model from language, recognise when the model needs adjustment and use evidence to guide the next move.
Cross-curricular practice prevents the skill from becoming trapped under the label English comprehension. A strong reading strategy should travel with the reader.
Primary and PSLE application
Primary learners can predict from titles, pictures, first lines and simple character goals, always adding because and revisiting the prediction.
At this stage, oral explanation is especially useful because it reveals whether the child genuinely understands why the strategy works. The teacher can then reduce prompts gradually until the learner initiates the move independently.
Secondary and SEC application
Secondary learners should use genre, structure, argument and causal chains while learning to keep confidence proportional to evidence.
Secondary texts contain longer sentences, implicit relationships, competing viewpoints and denser information. Practice should therefore include stories, articles, explanations and source-based texts rather than a single question style.
JC, IGCSE and advanced application
Advanced readers can predict rhetorical moves, evidence needs and counterarguments while remaining alert to deliberate surprise and qualification.
At advanced levels, readers also need calibration. They should know not only what interpretation is plausible but how strongly the evidence supports it, what alternative remains possible and whether the strategy is still worth the time it consumes.
Teacher diagnosis
A short think-aloud can reveal more than another full paper. Ask the learner to read a difficult span and verbalise what is noticed, what is expected, what is confusing and what action is chosen. The teacher can then distinguish knowledge gaps from strategy gaps, and concept problems from speed problems.
Feedback should preserve agency. Instead of supplying the answer, point to the process: “What changed in the third sentence?” “Which phrase supports that?” “What would make you revise this?” The goal is for the learner eventually to ask these questions internally.
Parent support
At home, keep prompts light. Ask the child to explain one difficult paragraph, show the clue that changed an interpretation, or describe what they did when the text stopped making sense. Avoid turning every reading session into a test. The purpose is to make strategic reading normal, not stressful.
Parents can also model uncertainty: “I thought this at first, but the next sentence changed my mind.” This shows that revision is a normal part of reading rather than evidence of failure.
Exam-day use
Under time pressure, use the smallest effective version of the strategy. Protect the evidence boundary, but do not overwork low-value ambiguity. Mark and return when necessary. Verification should be targeted to the line range, paragraph or relationship that controls the answer.
A useful stopping rule is: once the interpretation fits the relevant evidence, matches the question’s scope and does not require an extra unsupported story, answer and move on.
Measuring progress
Measure accuracy, independence, transfer and speed. A learner who becomes faster but less accurate has not improved in the intended way. A learner who explains the strategy perfectly but never selects it independently also remains incomplete.
Useful indicators include fewer full rereads, more precise evidence selection, quicker recovery from confusion, better paragraph gists, fewer unsupported claims and more consistent performance on unfamiliar passages.
Frequently asked questions
Do good predictions have to be correct?
No. They need to be reasonable when made and revisable when evidence changes.
Is prediction the same as inference?
Prediction projects forward; inference derives unstated meaning from current evidence.
Should I predict before every paragraph?
No. Use it where anticipation improves attention or structure.
Can headings mislead?
Yes. Treat them as clues, not guarantees.
Why keep wrong predictions?
They show how your model changed and which evidence mattered.
How long should prediction take in an exam?
Usually only a few seconds unless the task specifically requires it.
Research and reference points
Reading Rockets’ First Lines and Directed Reading Thinking Activity both make prediction, verification and revision explicit parts of comprehension.
Continue through SETC
Continue with English Reading and Comprehension, Making Inferences and English Learning Library.
The mastery target is a reader who can recognise when prediction is useful, execute it with evidence, revise when necessary and carry the same control into unfamiliar texts without waiting for a teacher to name the strategy.