Did you know that a reading log can be full of words and still reveal very little about what a child understood? A pupil may faithfully write every event in a chapter while never explaining why a character made a choice, which clue changed their prediction, or which sentence remains puzzling. That is a retelling, not yet a record of reading thinking.
A useful reading log is a short record of the reader’s response to a particular part of a text: what the text actually says, what the reader thinks it means, and why. It may include a brief summary, but it adds an evidence-based inference, question, prediction or clarification. The immediate improvement is simple: replace ‘Then this happened, then that happened’ with ‘I noticed this detail; it makes me think this; here is what I will check next.’
First, distinguish a plot summary from a response
A summary answers ‘What happened?’ It is a valuable skill, especially when the child has lost track of the story. But a response also answers ‘What did you notice or work out while reading?’ This does not mean an emotional reaction such as ‘It was nice’ is wrong. It means that an unexplained feeling cannot show whether the learner understood the passage.
For example, ‘The main character left the school hall and found a key’ recounts events. ‘The character hid the key even though the teacher was nearby; this may show she fears getting into trouble’ proposes an interpretation and points to a specific action. A reader can challenge or revise the inference if later evidence changes it. That is exactly the kind of thinking a learning log can make visible.
A complete worked example with an invented mini-story
Consider this original fictional passage created for this lesson, not drawn from a published book or an actual pupil’s work: ‘Mira reached the empty garden bench before the bell. There was a folded note under the seat. She recognised the handwriting, but instead of opening the note she tucked it into her bag. When her friend asked why she was quiet, Mira said she had forgotten something at home.’
A basic retelling would read: ‘Mira went to a bench, found a note, put it in her bag and told her friend she forgot something.’ That is accurate, but it offers no reason to return to the passage. A stronger log uses the detail that Mira recognises the handwriting yet does not read the note. It distinguishes what the narrator states from what the reader infers.
- Text detail: Mira recognised the handwriting but did not open the note.
- Inference: she may already suspect who sent the note and feel uneasy about its contents.
- Alternative explanation: she could be saving it for a private moment rather than feeling afraid.
- Question: What has happened between Mira and the person who wrote it?
- Check: read the next scene for a reaction to the note; do not call the suspicion a fact yet.
A model pupil response could be: ‘I noticed Mira knew the handwriting yet put away the note unopened. I think she might be worried about who wrote it, though she may simply want privacy. I will look for her reaction when she finally reads it.’ This is a few sentences rather than a page, but it shows observation, inference, uncertainty and a plan to test an idea.
Why evidence must be small and exact
Children sometimes choose a dramatic conclusion and then search for any nearby sentence to support it. If the log says ‘Mira hates her friend’, the invented passage does not establish that. The friend merely asks a question. A safer inference names the uncertainty and ties it to the action actually described. Ask: ‘Which word, action or change in the passage led you there?’
A reading log is not improved by copying a whole paragraph. A short accurate reference—such as ‘recognised the handwriting’—may be enough when the teacher permits it. Keep page numbers, chapter headings or digital locations consistent with the edition being used, because classmates may read different editions. Explain the evidence in the learner’s own words unless a quotation serves a particular purpose.
Four kinds of useful entries, not four compulsory boxes
- Inference: ‘The character avoided answering. I think she is hiding something because…’
- Question: ‘Why did the narrator describe the locked gate twice? I need to watch what changes.’
- Revision: ‘I predicted the neighbour was angry, but the later apology suggests I misread the silence.’
- Clarification: ‘I confused two speakers. I reread the dialogue tags and now know who said what.’
These are choices, not a checklist that every entry must fill. Some pages are best summarised; others are more valuable for vocabulary, setting, motive or a surprising change. The purpose of the lesson and the demands of the text decide which response is most useful. A child who is still decoding the sentences may first need oral retelling and targeted word support.
What happens when the child’s inference is wrong?
This is where the log can become unusually helpful. Imagine the next fictional chapter reveals that Mira wanted to open the note together with her sister because it concerned their family. The earlier ‘she may be worried’ inference was reasonable but incomplete. Rather than crossing it out as a failure, the learner can write: ‘I thought she feared the sender. New evidence suggests privacy and family cooperation mattered more.’
The evidence-supported revision shows understanding growing over time. The best readers do not merely collect correct guesses; they notice when new sentences overturn an earlier interpretation. A teacher should reward the honest revision and ask which detail caused it.
A five-minute routine for home or class
Invite the child to choose one moment from ten or fifteen minutes of reading. The log does not have to record every page. Let them state the scene in one sentence, select one detail, say what that detail suggests, and identify a question or check. If the learner gives a sweeping claim, ask for the exact passage that supports it. If the learner gives a perfect quote but no thinking, ask what the chosen line changes about their interpretation.
For reluctant writers, let the first response be spoken. The parent might say, ‘Tell me one thing you now suspect, and show me the clue.’ Once the child can say a supported idea, transfer it into two or three sentences. This is better than requiring a long response that the pupil cannot yet formulate.
How teachers can model the thought process without supplying every answer
A teacher can demonstrate an ordinary reading decision aloud: ‘I thought the character was calm. But the sentence about her trembling hands makes me reconsider. Maybe the narrator does not say how she feels directly.’ The child hears the movement from clue to interpretation rather than a mysterious correct answer arriving fully formed.
Then try shared practice on a new paragraph. Ask the class to generate two possible inferences, check which details support each and reject one that lacks evidence. After a few examples, stop giving the opening words and let the pupil find the clue independently. Guided work should be a bridge, not a permanent script.
The Education Endowment Foundation’s reading comprehension overview describes questioning, clarifying, inferring, summarising and monitoring as teachable approaches. It also notes that evidence varies by context, so a reading log should support actual reading and discussion rather than become a guaranteed recipe for higher scores.
A diagnostic that tells you which skill needs repair
- Can the child retell the basic event accurately? If not, work first on who did what and when.
- Can the child distinguish the passage’s stated action from a guess about motive? If not, practise ‘the text says / I infer’.
- Can the child locate a relevant clue? If not, practise rereading one short scene and choosing a precise phrase.
- Can the child explain an alternative interpretation? If not, compare two plausible ideas against the evidence.
- Can the child revise after new evidence? If not, revisit a previous entry and model what changed.
The same weak log can have different causes. A child who misunderstands the story needs comprehension support; one who understands orally but struggles to write may need sentence rehearsal; one who copies everything may not know how to choose an important moment. Diagnosis keeps the help targeted.
Progress from guided to independent responses
At first, give two short possible responses and ask which makes a claim supported by the text. Then ask the learner to complete one evidence-and-inference pair. Next, remove the prompt and ask for a response to a fresh story excerpt. Several days later, use a different genre such as an information text, and replace ‘character motive’ with ‘author’s explanation’ or ‘unanswered question’.
A parent or teacher can record whether the learner selects a relevant clue without help, uses cautious language when uncertain, and revisits a question after reading further. Those are observable behaviours. A higher word count, attractive stationery or more entries per week does not automatically indicate deeper comprehension.
A second worked task: an information text
Suppose a fictional article says that a school garden’s shaded soil stayed moist for longer than the sunlit soil during three observations, but the two areas had different plant cover. A mere retelling says ‘The shaded soil was wetter.’ A stronger log says, ‘The result might be connected to shade, but plant cover also differed; I need more evidence before claiming shade alone caused it.’
The lesson transfers beyond novels because good readers distinguish observation, causal explanation and unresolved questions. The student can learn to ask ‘What evidence is supplied?’ in a science passage as well as ‘Why did she act like that?’ in fiction.
Common mistakes that look like effort
One mistake is turning each log into a book review: ‘The chapter was amazing and exciting.’ The reaction can be a starting point, but ask what detail produced it. Another is asking only factual recall questions; these can help check understanding, yet a useful log sometimes needs an inference or an uncertainty rather than a quiz.
A third mistake is copying a model response too closely. The pupil may learn the vocabulary of reflection without making a fresh reading decision. Use a new passage and ask for a new clue. A fourth mistake is making the log punitive: if every page requires a long entry, the child may begin reading only to fill a form. Keep the task proportionate to the reading goal.
Limitations and appropriate expectations
Not every book is suitable for the same response. Poetry, graphic narratives, news items and beginner decodable stories need different prompts. When comprehension is constrained by unfamiliar language or insufficient background knowledge, explicitly teach what the text means instead of merely demanding a deeper inference.
Readers may legitimately disagree about an ambiguous character motive. The goal is not uniform opinions, but a sensible claim with relevant evidence and a willingness to change when warranted. School marking instructions may demand a particular response format; follow those while preserving the distinction between retelling and interpreting.
Frequently asked questions
Is a summary always bad in a reading journal?
No. Summarising is essential when a reader needs to keep the main events straight. The point is to add a purposeful thought when the activity is assessing comprehension beyond recall.
How long should a reading log be?
Long enough to contain a specific clue and an explained response. Two or three well-chosen sentences can be more revealing than a page of chronology.
Can a child write a prediction that turns out wrong?
Yes. Encourage them to name the clue behind the prediction and revisit it after new evidence appears.
What if my child cannot find a quote?
Start by locating an action or paraphrasing a scene accurately, then teach how to use a small precise excerpt where appropriate.
Should parents correct the opinion?
Ask for the evidence and clarify factual misunderstandings. Do not replace every defensible interpretation with an adult’s preferred answer.
Continue learning with eduKate
Read the existing Core Aim guide on supporting a character interpretation with two moments and its companion on setting versus mood. Explore the eduKate Education Hub, and see the Clementi tutorial method for the wider approach of diagnosis, guided explanation and independent checking.
A useful reading log need not recount the whole journey. It should preserve the moment when a reader noticed something, made a reasoned claim and learned what to look for next.