"The Answer Is Not Inside!"
Somewhere in P3, and unmistakably by P4, English comprehension changes its rules without telling anyone.
My daughter came to me with a passage about a boy named Wei Jie who "stared at his shoes and mumbled a reply" when his teacher asked about his homework. The question: How do you think Wei Jie felt? Give a reason for your answer. She had read the passage three times. She was close to tears. "Mummy, the answer is not inside!"
She was right, in a way. For three years, comprehension had been a finding game. Where did the family go? What did Grandma cook? The answer sat in the passage like a fishball in soup — you just had to spot it and scoop it out. Nobody had told her the game had changed. Now the paper wanted answers the writer never wrote, and she felt, quite reasonably, that this was cheating.
It is not cheating. It is inference — and the jump from literal to inferential comprehension around P3 and P4 is one of the biggest invisible walls in primary school English. Invisible, because no letter comes home announcing it. One of the biggest, because inference never goes away: it grows through upper primary, dominates comprehension at P5 and P6, and, frankly, runs most of adult life.
The good news: inference is teachable, at home, mostly through conversation, and it can be genuinely fun. Here is what has worked with my own kids and what I have seen work with others.
What Inference Actually Is
Strip away the jargon and inference is detective work with two ingredients:
- Evidence — the actual words on the page. Wei Jie stared at his shoes. He mumbled.
- A clue from life — what those words usually mean in the real world. People who stare at their shoes and mumble are usually embarrassed, guilty, or ashamed. Proud people look up and speak clearly.
When I explained it this way — "the writer gives you the clues; you're the detective; the answer is what the clues add up to" — something unlocked. The complaint "the answer is not inside" became "okay, so what are the clues?" That reframe, from finding to solving, is honestly half the battle. Kids love being detectives. They hate feeling tricked.
Learn to Spot the Question Type First
Before answering anything, a P4 child should learn to sort the question: is this a FIND question or a THINK question? Different questions, different tools.
Stems that signal a THINK (inference) question:
- Why do you think… — the words "do you think" are the giveaway; the passage will not say it directly
- How did X feel when… — feelings are almost never stated outright in good writing
- What does this tell us about… — "tell us about" means add the clues up
- What is likely to happen next? — prediction is inference pointed forwards
- Which word tells you that… — this one is inference in reverse: the conclusion is given, hunt the evidence
- What did X do when…
- Where did the family…
- Who gave…
- When did…
The Evidence-Plus-Clue Habit
Here is the core habit, the one thing I would teach if I could only teach one: every inference answer has two parts — point to the words, then say what they suggest.
We use a sentence frame, said aloud until it became automatic:
"I think ___ because the passage says ___, and that usually means ___."
For Wei Jie: "I think he felt embarrassed, because the passage says he stared at his shoes and mumbled, and people usually do that when they feel bad about something."
With younger or more reluctant kids, I make it physical — the two-finger rule. One finger presses on the evidence in the passage. The other finger taps their own head: what do I know about what this usually means? No answer leaves their mouth until both fingers have done their job. It looks silly. It works, because it makes the invisible structure of a good answer visible and touchable.
Notice what the frame quietly prevents: it makes a bare "He felt sad" impossible, because the frame demands a because. And it makes wild guessing impossible, because the frame demands actual words from the passage. Both traps, blocked by one habit.
Storybooks Are Your Practice Papers
You do not need assessment books to practise inference. You need bedtime stories and the willingness to interrupt them.
The pause-and-ask read-aloud is the gentlest inference training there is. Reading together, you stop at a juicy moment and wonder aloud:
- "Hmm, why do you think she hid the letter?"
- "How is he feeling right now? What tells you?"
- "What do you think happens when Mum finds out?" — then turn the page and check.
Two notes from experience. First, picture books are not babyish for this — even at P4. Faces, body language, what is drawn in the background: illustrators bury inferences in pictures, and "how does the fox feel — look at his ears" is world-class training. Second, use books your child already loves. Inference practice on a beloved series feels like gossiping about mutual friends. Inference practice on a worthy-but-dull passage feels like homework, because it is.
The Wet Umbrella Game
The real breakthrough in our house did not come from a book at all. It came from the umbrella.
We walked in one evening and there was a dripping umbrella by our neighbour's door. I said, without thinking, "So — what can we tell?" My son looked at it, looked at me, and said "Uncle came home already. And it rained where he was, even though never rain here." Reader, that is an inference with evidence, produced joyfully, by a child who thirty minutes earlier would have told you he "cannot do comprehension."
"What can we tell?" became our family catchphrase, and Singapore is generous with material:
- The auntie in the lift with six red plastic bags — where has she been? What might her family be doing tonight?
- The long queue at one chicken rice stall while the next stall stands empty — what can we tell?
- Big brother goes into his room and the door closes a little too hard — what can we tell, and should we ask him about it now or later?
- Half the void deck chairs stacked and bunting going up — what is coming this weekend?
- Someone on the MRT holding balloons and a cake box — where are they going? What is the evidence?
The Five Traps That Eat Marks
Now for the honest section: the specific ways bright, hardworking P4 kids lose inference marks. I have watched every one of these happen at our dining table.
Trap 1: Copying whole sentences
The child senses the answer is "somewhere around there" and copies the entire sentence, sometimes two, hoping the right words are inside. Teachers call it lifting. The problem: copying proves finding, not understanding — and for a THINK question, the copied sentence usually contains the evidence but not the conclusion, so the answer is incomplete at best.
The fix we use: answer with the book closed, in your own words, out loud first. Then open the book and borrow only the key words you need for the evidence half. Own words for the thinking, passage words for the proof.
Trap 2: Feelings without evidence
How did Wei Jie feel? — "He felt sad." Full stop. Even when the feeling is right, half the answer is missing, because the question almost always wants the reason. The evidence-plus-clue frame exists precisely to make this mistake structurally impossible. If your child does one thing differently after this post, make it this: no feeling without a because.
Trap 3: A three-word feelings vocabulary
Happy, sad, angry. Maybe scared on a good day. A child who owns only three feeling words will force every character into one of three boxes, and the marker is looking for embarrassed, disappointed, relieved, anxious, proud, guilty, jealous.
We keep a feelings list on the fridge and add to it from books — when a character is clearly humiliated rather than merely sad, that word goes on the list. Then we use them in real life: "You look disappointed, not angry — am I right?" A child who has felt the difference between disappointed and angry can read it in a character.
Trap 4: Inference without a tether
The opposite failure: some children, once given permission to go beyond the text, go far beyond it. Wei Jie mumbled at his shoes — "maybe an alien stole his homework and he cannot say!" Creative, zero marks. Inference is imagination on a leash: every conclusion must be tethered to actual words in the passage. The test I use at home: "Show me the words that made you think that." No words, no answer — try again.
Trap 5: Grabbing the first mention
The question asks about the ending; the child scans, finds the character's name in paragraph one, and answers from there. Speed-reading habits from the literal years die hard. The fix is boring but effective: for THINK questions, locate the moment the question is about first, re-read those two or three lines properly, then answer.
Marking at Home Without Tears
A word about tone, because inference practice can turn sour fast if every wrong answer gets a red pen.
When my daughter gives a wrong or half answer, I try not to say "wrong." I say, "Show me the words that told you." Sometimes, walking me through it, she catches the mistake herself — which is worth ten of my corrections. Sometimes her evidence is genuinely reasonable and the marker's answer is just more reasonable — and I say so, because a child whose sensible-but-imperfect reasoning gets respected will keep reasoning, and a child who gets "wrong, the answer is B" will go back to guessing.
Little and often beats long and rare. One passage done slowly and talked about beats five passages rushed and marked in silence. We keep a rhythm of short daily practice — some days that is ten minutes of questions on BrightBit after dinner, some days it is just the wet umbrella game on the walk home — and save full papers for occasional weekends.
Bigger Than the Exam
I will end with the thing I actually believe. Inference is not a comprehension skill that happens to be useful in life; it is a life skill that happens to be tested in comprehension.
A child who can read "stared at his shoes and mumbled" and conclude embarrassed is the same child who can look at a classmate sitting alone at recess and conclude maybe he wants company but doesn't know how to ask. Evidence, plus what I know about people, equals understanding — and then, if we have raised them well, equals kindness.
That is what we are really teaching when we teach reading between the lines. The marks are just the receipt.