Why practising inference doesn’t work

//Why practising inference doesn’t work

In my last post I argued that thinking about English as a ‘skills based’ subject is counter-productive. One response to this was to say, “Hang on, what about practice. If you can practise something you become more skilled at it, so how can you say English isn’t a skills-based subject?”

It seems obvious that “just knowing” something is different from practising it. Pretty much anything we do can be improved through practice. But, the role of practice changes depending on whether you think English is a skills based subject or not.

In the skills-based approach it makes sense to practise the skill of inference because by practising it you’ll become better at it. My view is that there is no such thing as a skill of inference, and so, if it doesn’t exist, you can’t practice it. To be clear, I’m not saying inference doesn’t exist, – obviously you can tell students what an inference is and show them lots of examples –  just that it’s not a skill that can be practised. Let’s see if I can explain why.

An inference is defined as “a conclusion reached on the basis of evidence and reasoning”. So, to think about whether we can practise making inferences we need to consider whether by drawing conclusions based on evidence and reason we’ll become better at drawing conclusions based on evidence on reason. This sounds logical, but we need to consider what we are actually practising. How do you actually draw a conclusion from evidence and reason? Well, first you need to weigh the specific evidence. How do you know what the evidence is telling you? How do you know the importance of the evidence? Then, how do we apply reason to this evidence? Are these generalities that can be learned, or do they depend on specific instances?

Robert Marzano suggests that we can teach inference by posing four questions to students:

  • What is my inference?
    This question helps students become aware that they may have just made an inference by filling in information that wasn’t directly presented.
  • What information did I use to make this inference?
    It’s important for students to understand the various types of information they use to make inferences. This may include information presented in the text, or it may be background knowledge that a student brings to the learning setting.
  • How good was my thinking?
    According to Marzano, once students have identified the premises on which they’ve based their inferences, they can engage in the most powerful part of the process — examining the validity of their thinking.
  • Do I need to change my thinking?
    The final step in the process is for students to consider possible changes in their thinking. The point here is not to invalidate students’ original inferences, but rather to help them develop the habit of continually updating their thinking as they gather new information.

I’m going to assume that readers will consider themselves reasonably practised at making inferences. Let’s try applying these four questions to a short passage from Finnegan’s Wake:

riverrun, past Eve and Adam’s, from swerve of shore to bend of bay, brings us by a commodius vicus of recirculation back to Howth Castle and Environs. Sir Tristram, violer d’amores, fr’over the short sea, had passencore rearrived from North Armorica on this side the scraggy isthmus of Europe Minor to wielderfight his penisolate war: nor had topsawyer’s rocks by the stream Oconee exaggerated themselse to Laurens County’s gorgios while they went doublin their mumper all the time: nor avoice from afire bellowsed mishe mishe totauftauf thuartpeatrick: not yet, though venissoon after, had a kidscad buttended a bland old isaac: not yet, though all’s fair in vanessy, were sosie sesthers wroth with twone nathandjoe. Rot a peck of pa’s malt had Jhem or Shen brewed by arclight and rory end to the regginbrow was to be seen ringsome on the aquaface. The fall

The rest of the section is here.

  • What is your inference?
  • What did you use to make that inference?
  • How good was you thinking?
  • Do you need to change your thinking?

Was your skill in making inferences any help? Did the questions help you in your thinking?

  • My inference: I have a general sense that some sort of route or journey – possibly involving Howth Castle – is being described.
  • What I used: “past”, “from swerve of shore to bend of bay”, “brings us” all suggest a journey is being undertaken or directions given. References to “North Armorica” and “Europe Minor” make me think of old maps – possibly someone has travelled over the Atlantic?
  • How good was my thinking? Honestly, I have no idea. I could be completely on the wrong track.
  • Do I need to change my thinking? How would I know? I have no idea if I’m even vaguely right.

If I wanted to improve my ability to make an inference I would need to know more about what Joyce was trying to achieve. Have a look at what Cliff’s Notes has to say and see if you could have inferred any of this without knowing an awful lot.

Once you know that inferences can be made about a text, then you can start trying to make them, but unless you know something you won’t be able to make a worthwhile inference. So what does it mean to be skilled at making inferences? My argument is that this is indistinguishable form being knowledgable.

Practising the wrong things leads to Cargo Cult Teaching. Students know what an essay is supposed to look like, they know the sorts of words and phrases they should use, but they have no understanding of the underlying content and therefore whatever they say is likely to be superficial at best and fatuous at worst.

What you can practise is retrieving and applying what you know. This will be the subject of my next post.

You should also read Willingham & Lovette’s article, Can Reading Comprehension Be Taught?



  1. chrismwparsons April 29, 2018 at 4:58 pm - Reply

    Now, I’m going to make a comment here because – over time – I have become skilled (ok – comfortable with) making comments on your blogs. I get your post, and I agree with it. What to I infer from it…(in terms of seeing the implications)?

    Well, I haven’t even tried to make use of the algorithm. However, for what it’s worth, my personal belief is that inference is simply an uncontrollable act of association – the brain does it in spite of us – and that the best which deliberate strategies can do is simply to remind us to pay some attention. Now – I’m wondering how often I call to mind a checklist of things to do when I start to make inferences about a situation…? Hmm…

    Perhaps I simply need more practice at using the checklist…? Would that make it habitual over time and result in me using it without a prompt, or would I just find myself going ahead and either making inferences or…er….not?

    • David Didau April 29, 2018 at 5:47 pm - Reply

      Well, taking the checklist example in the post, what use is it if you can’t make any worthwhile associations?

  2. Rebecca Tulloch April 29, 2018 at 5:45 pm - Reply

    Aren’t we practicing reading? Isn’t getting better at reading that helps? The more you read the more knowledge you have of narrative structures that writers can use or tricks they can play. The more you read the more skillful you get at reading inferences. The skill and the knowledge are as important as each other.
    My son is superb at understanding the inferences in complex texts like Pax and Wonder that we read at bedtime, but often fails in the short inference tasks his teacher sets. I think he is brilliant at it, she does not. He will have to learn to play the education game of inference, whilst being quite skillful as a reader at home.

    • David Didau April 29, 2018 at 5:52 pm - Reply

      Practising reading is worse than practising inference. What does it even mean? Does reading 100 YA novels provide the ‘skill’ necessary to read a Dickens novel? You become more skilful at making inferences because by reading we are exposed to more information about the world. Your son knows enough about the world to get the associations required in YA novels. That’s great. These novels work because enough young people have enough knowledge to enjoy reading them independently. The “education game” of inference is a canard. It’s meaningless. Whether one can do these exercises depends on what they know in relation to the task and whether they can replicate the structures in which the teacher decides the inference must be made.

      • mariesnyder April 29, 2018 at 7:11 pm - Reply

        What about practicing reading as analogous to practicing weight-lifting? If you only ever lift ten pound weights (read Y.A. novels), you won’t much improve. So you have to keep lifting the weights that are just a bit above your ability to manage easily. You should find it a bit of a struggle. Likewise, we have students read books that are individually chosen to be just above the types of books they’re currently reading. Then a bit more and a bit more will improve their reading ability. And then the more we do that, and the more content we get from various sources, the more we can understand connections between texts and understand what authors want us to infer. I think practicing reading as a skill goes hand in hand with learning the content and form of various texts as well, though. You can’t have one isolated from the other. That would be like lifting increasing heavy weights without eating enough food to sustain yourself. Something like that!

        • David Didau April 29, 2018 at 9:08 pm - Reply

          Well, I know what you mean, but I think the mistake is in seeing reading as a homogenous skill. The bit that improves through reading is knowledge. Knowledge of the world, of stories, of literary conventions, of vocabulary etc. The mechanical aspect of reading – decoding – improves through practice of phoneme blending and increasingly fluent recognition of phoneme/grapheme relationships. Although reading ‘in the wild’ can offer some useful practice, there are better ways to increase fluency. This is important because those 20% of children who leave primary school unable to read fluently enough to access an academic curriculum will never read for pleasure because it’s too much like hard work. They require targetted, specific interventions to practice acquiring those aspects of knowledge required for fluency.

        • Michael Pye April 30, 2018 at 4:13 pm - Reply

          Weightlifting programming is an incredibility complex and controversial topic. Basic progressive resistance (which is what you describe as adding a few pounds) stops working after about a year. Emphasis on technical points, scaling back and re-increasing, rep ranges, variant exercises etc: are all used to emphasize individual aspects before reintegrating. Expert levels in weightlifting, Olympic weightlifting, Crossfit, Kettle Bell Sport etc are all predicated on large amounts of theoretical and procedural knowledge (and much controversy).

          Staying in shape a bit and not losing muscle mass doesn’t require any of this: In the same way most recreational reading doesn’t. Of course we want to stretch peoples level of reading not simply reinforce their current competency. Incidentally high level athletes lose lots of their physical advantages over time but they maintain the technical knowledge for a lifetime.

          • Michael Pye April 30, 2018 at 4:20 pm

            Upon re-reading I don’t think I made a clear point. Without domain specific strategies stretching, practicing and reinforcing a higher standard becomes increasingly more difficult and demoralizing. New Knowledge however can however still be easily acquired and these new facts give us plenty of material to practice inference and other skills as a logical extension of ordering and analysing these new ideas.

  3. chrismwparsons April 29, 2018 at 7:33 pm - Reply

    In my earlier response above I was being a bit playful – I do agree with you that the underlying knowledge is necessary, and – almost – sufficient. I do think that the vast majority of inference takes place automatically whilst our conscious (type 2) thinking brain sits by arrogantly and takes the credit for it.

    However, we all have experience of someone who appears to us as being – in the moment – surpristingly ‘dense’ – in other words, they are people who fully have the knowledge to understand something, but are missing the point. In this sense, I think there is scope for some kind of simple technique – teachable – which could help the penny drop a little…

    Personally, I …wait for it… take a deep inhalation…. would say that the left brain/right brain distinctions between close-up, detailed logical analysis versus wider, more associative, big-picture perception give a simple picture of where people tend to get stuck here. Consequently, I recommend a very simple, very easy to grasp for the modern generation, technique of simply……………… “zooming in and out until you get it”.

    Yeah – if they haven’t got the knowledge it just ain’t going to jump out. But, we’ve got to accept that even if they have got the knowledge it isn’t guaranteed to jump out either……

    • David Didau April 29, 2018 at 9:10 pm - Reply

      You know I invented zooming in and out, right? I came up with it years ago and first wrote about it 2011 https://www.learningspy.co.uk/english-gcse/zooming-in-and-out/ 😉

      • chrismwparsons April 29, 2018 at 9:26 pm - Reply

        Now, you’re never going to believe this, but I don’t think that I’ve ever read that, nor anyone else mentioning it either! But then again, I don’t recall an awful lot regarding things which I think seem original in my mind, so I’m more than happy to bow down to you as the originator of that idea!

        Irrespective, would I not be reasonable in believing that the principle is about the limit regarding worthily teachable inferential heuristics…?

        • David Didau April 30, 2018 at 6:06 pm - Reply

          As Willingham argues, it’s definitely worth telling students about these heuristics, but there’s very little point in practising them: http://www.danielwillingham.com/uploads/5/0/0/7/5007325/willingham&lovette_2014_can_reading_comprehension_be_taught_.pdf

          • chrismwparsons April 30, 2018 at 6:10 pm

            Ah – but if we’ve simply got the heuristic as a declarative idea in our long term memory, then we’re not likely to make much use of it (and it would be only a type 2 – slow resource). Surely the point is that it needs to be a reflexive habit which we find ourselves naturally resorting to without prompting or delay…? Doesn’t this require deliberate practice in the first instance?

          • David Didau April 30, 2018 at 9:05 pm

            I don’t think so. Back when I was at school English consisted filet of reading books and writing stories. At no point was I ever told about making inferences or analysis or any of the other ‘skills’ which are now the bedrock of most English lessons. I just sort of picked it up as a fairly natural consequence of reading.

          • chrismwparsons April 30, 2018 at 10:18 pm

            I get what you’re saying – I think it was the same for me, but I’m wondering whether there might be the (hastily made-up) “I didn’t need it, so it can’t help,” or the “I didn’t feel the benefit so there’s no benefit to be felt,” fallacy…

          • David Didau May 1, 2018 at 9:39 pm

            Ok. So, I didn’t need to be taught to make inferences to make inferences and teaching inference to my students made no difference to their ability to make inferences. What else is there?

  4. Chester Draws April 30, 2018 at 5:36 am - Reply

    “How good was your thinking”

    It’s the best I got!

    I find these sorts of questions patronising, when not utterly useless. They suggest that people are knowingly not thinking at 100%.

    When I don’t understand something, telling me to try to understand it better is useless. I would if I could.

    I refuse, point blank, to ask my students to evaluate the level of their thinking. It acts only to demoralise. Not only are they failing to understand, they are now being, effectively, taunted about.

  5. Michael Pye April 30, 2018 at 4:15 pm - Reply

    Reminds me of practicing Akido. I was always told to relax which of course had the opposite effect. I just wanted clear feedback on what to do instead.

  6. […] composed on many thousands of individual components of knowledge organised together as schema. In my last post I tried to demonstrate that practising ‘inference skills’ won’t actually help […]

  7. […] they do not improve through practise. Practising essaying writing leads to cargo cult essays. Practising inference is a dead-end that not only wastes children’s time but sucks all the joy out of […]

Constructive feedback is always appreciated

%d bloggers like this: