Inference Is Not a Skill (and Why I Don’t Teach It)

Every minute that parents and teachers spend trying to teach inference as a discrete skill could be spent reading quality books.

That is the basic argument of this post. I have tried to teach inference as a discrete skill and have, in the past, developed entire schemes of work where I attempted to improve my pupils’ inference by getting them to practise so-called “inference skills” in isolation. I do not do this any more, and I would strongly encourage those who are dedicating significant chunks of time to teaching inference in isolation to stop doing so.

Nowadays, I teach a Year 6 class two days a week and tutor privately for the rest of the week, mostly for the 11+. Across a whole school year or a full 11+ preparation cycle, I probably spend a grand total of about one hour talking about inference.

All I do regarding inference is:
• explain what inference is
• show children what inference questions look like
• model how to answer them

Then I stop. I do not run 11+ “inference courses”. I do not have folders of “inference worksheets” with a clip-art detective holding a magnifying glass. In my opinion, people who try to teach inference in this way are wasting their own time and their child’s or pupils’.

What is inference?

Inference is simply the ability to read between the lines and pick up on subtle clues. That is it.

A child reads a passage, connects it to what they already know about language and about people, and comes to a sensible conclusion that is not directly stated.

For example, if a story says:
“Tom stared at the floor, chewing his lip. ‘It was fine,’ he muttered.”

Most adults and confident readers will infer that it was not fine at all. We notice the body language, the muttering, the lack of eye contact. Nobody has to spell it out for us.

That jump, from what is written to what is really going on, feels like a “skill”. It is not. It is the result of lots of other things working well together.

Inference is impossible without knowledge

This is the part that people struggle to understand, and for teachers, it is a hangover of the 1990s and 2000s when engagement, edutainment and “skills” were in vogue while knowledge was treated as a dirty word.

You cannot read a piece of text and make an inference if you do not understand the words, the situation or how people behave. If any of those parts are missing, inference collapses. When a child struggles with an inference question, it is never because they are “not good at inference”. It is always because there is a gap underneath it.

A vocabulary gap

They do not understand key words or phrases in the sentence.
If a character gives a “wry smile” or someone speaks in a “curt” tone, and the child does not know what those words mean, they will not pick up the emotional clue. This is not an inference problem. It is a vocabulary problem.

A knowledge gap

They do not know enough about the world, history or the context of the story.
If a text mentions rationing, evacuation, a strike, or something as simple as a boarding school or a long-distance train journey, some children have very little to connect that to. They are trying to infer in a vacuum.

An experience or behaviour gap

They have not seen enough of how people act, or they have not had enough conversations about feelings and motives.
If a character “shrugs and changes the subject”, many adults recognise that as embarrassment, avoidance or guilt. Some children have not yet seen that pattern. Again, that is not an inference problem. It is a lack of experience and discussion.

When you fix those gaps, the “inference problem” usually disappears.

Inference is not a skill

Because of all this, I do not believe inference exists as an isolated skill that can be developed in isolation. This is in contrast to retrieval (finding and recalling key information from a text) and fluency (reading smoothly with accuracy, pace and expression), which can be taught, practised and developed over time.

You cannot take a child who does not understand half of what they read, give them twenty inference questions a week, and expect anything meaningful to change. They may get better at spotting what the question is looking for, but the underlying ability to read between the lines does not suddenly grow because they did an inference task about a different text or topic last week.

Inference is better thought of as a concept and an outcome, not a skill. It is something that happens naturally when the right foundations are in place. It depends on:
• strong vocabulary
• secure background knowledge
• an understanding of people and situations
• an understanding of how written language works

If those foundations are weak, practising inference will not fix it.

Opportunity Cost

Teachers, tutors and parents all face the same problem: limited time.

Every choice has an opportunity cost. If you spend half an hour on inference worksheets, that is half an hour you cannot spend on something else. I want to be clear that I am not denying the existence of inference and I do talk about inference with my pupils. I just do not teach inference as a standalone skill. Once I have explained what inference is and shown what those questions look like in papers such as the SATs or the CSSE 11+, I see very little value in targeted inference practice. Every minute spent “doing inference” in isolation would be better used elsewhere.

For example:
• reading a high-quality book together
• exploring vocabulary deeply and discussing an author’s language choices
• unpicking why a character behaved in a particular way
• making links to history, geography, culture, behaviour, their own experiences and other books or texts they have read

Those are the things that actually improve a child’s ability to infer.

So what should we do instead?

If we accept that inference depends on knowledge and experiences, the solution becomes much simpler.

Children need more of the things that feed inference:

Read quality books, regularly.
Not just extracts for comprehension, but whole stories, quality picture books, chapter books and novels. To get the best out of a book, it helps if you have read it yourself first. This lets you check that the content and themes are suitable, and it also means you can ask more effective questions. It gives you the background you need to talk about ideas, explain references, and build your child’s general knowledge by connecting the book to your own experiences, knowledge and understanding of the world.

Talk as you go.
Pause to explain words. Ask, “What do you think he is really feeling here?” or “Why do you think she did that?” without turning it into a test.

Build knowledge of the world.
If you are preparing for the CSSE 11+, do not overlook science and the non-core subjects. History, geography, science, museums, documentaries and conversations at the dinner table all give children the background knowledge they need to understand more complex texts.

Notice how people behave.
Talk about friendships, arguments, mistakes, kindness, jealousy. When they see similar patterns in books, they are much more likely to understand what is going on beneath the surface.

If you do those things, inference turns up as a by-product. Children start reading between the lines without needing to practise inference as a separate activity. As I write this (in November 2025), I am entering my tenth year of teaching in primary schools and my sixth year of tutoring privately. In that time, I have taught and tutored well over 300 pupils. I have never met a single pupil who could infer easily but rarely read quality books and had weak general knowledge, nor have I met a pupil who struggled to infer but read quality books regularly and had well-rounded general knowledge. I am not exaggerating here; not a single one.

The Heart of My Argument

• Inference is impossible without a good understanding of writing, the world and human behaviour.
• Inference is not a skill and cannot be developed in isolation.
• When children struggle with inference, the issue is always a gap in vocabulary, knowledge or experience.
• Beyond explaining what inference is and showing what those questions look like, time is far better spent reading and talking about books.

Reading makes you clever.