Theory Library Evidence Base Craik 1972

Thinking about meaning helps memory

Levels of Processing: A Framework for Memory Research

Craik, F. I. M., & Lockhart, R. S. · Journal of Verbal Learning and Verbal Behavior · 1972
Fox reading

In one line: how well we remember something depends a lot on how we thought about it at the time. Thinking about what it means usually beats noticing how it looks or sounds.

What they did

This isn’t an experiment. It’s an argument about how memory works, from two psychologists in Toronto, built on many earlier findings.

At the time, most psychologists pictured memory as a set of boxes (a brief sensory store, a short-term store, a long-term store), with information moving from one to the next. Fergus Craik and Robert Lockhart suggested a different way to look at it. What matters, they said, is how deeply you process something when you meet it. You can notice only its surface (what it looks like, what it sounds like) or you can think about what it means and how it links to what you already know. They saw this as a sliding scale, from surface to meaning, not a set of fixed steps.

Later experiments by Craik and Endel Tulving (1975) tested three points on that scale. People saw a list of words and answered a quick question about each one, such as:

  • How it looks: “Is the word in capital letters?”
  • How it sounds: “Does it rhyme with train?”
  • What it means: “Is it a type of animal?”

Then, without warning, they were asked to remember the words.

What they found

  • Meaning beats surface. In Craik and Tulving’s tests, people remembered far more of the words they’d thought about for meaning, though nobody had told them to try to remember.
  • Trying to remember isn’t the main thing. What people did with the words mattered more than whether they meant to learn them.
  • Going over something again adds little. Craik and Lockhart argued that just repeating something does little for later memory. Re-reading notes feels productive because it becomes familiar, but familiar isn’t the same as being able to recall it.
  • Meaning comes from links. Craik and Lockhart argued that deeper processing means connecting new material to what you already know.

How much should you trust it?

A useful, well-supported idea, but with two big catches.

  • The basic finding, that thinking about meaning usually beats noticing the surface, has been repeated many times in the lab, mostly with word lists.
  • “Deep” was never defined on its own terms. Critics pointed out that “deep” ends up meaning “whatever got remembered”, which goes round in a circle.
  • What helps depends on how the memory will be used. In one study, people who had thought about rhymes did better on a rhyming test than people who had thought about meaning (Morris, Bransford & Franks, 1977). So practise in the way you’ll need to use it.
  • It has hardly been tested in healthcare teaching.

If you need to convince someone

In our words (a paraphrase, not a quotation): people remember what they’ve thought about, not just what they’ve seen or heard. Ask them to work out what it means and what they’d do with it.

What it means for you

Start with what people need to do differently, then make them think about that, not just hear it.

  • Ask, don’t summarise. “So, the key point is…” lets people listen with half their mind. “What would this change on your shift?” makes them work out what it means.
  • Use a real case. A short scenario that asks a pharmacy assistant or a porter to apply the idea gets them thinking about meaning, which a bullet-point summary doesn’t.
  • Probe right answers. “Yes, exactly” ends the thinking. “Right, and why does that matter on a busy ward?” starts a second round.
  • Don’t rely on depth alone. Add recall. Pulling something out of memory strengthens it in a way that re-studying, even thoughtfully, does not, and researchers still debate exactly why. So a quick “what do you remember from last week?” tends to beat a recap slide. See retrieval practice.

How it appears in Teaching That Lands

It isn’t taught directly in the three core sessions. It sits in the background of elaboration, one of the four strategies participants sort in Session 2: linking new ideas to what you already know, or asking “where could this work?”

The small print

  • Two kinds of repeating: Craik and Lockhart separated simply repeating something to keep it in mind (maintenance rehearsal) from thinking about what it means (elaborative rehearsal). They argued the first does little for later memory.
  • Circularity: the criticism that depth can’t be measured apart from the memory it predicts came from Eysenck and Baddeley in 1978.
  • Transfer-appropriate processing: the idea that memory is best when the way you learn matches the way you’re tested (Morris, Bransford & Franks, 1977).
  • Retrieval practice: Karpicke and Blunt (2011) found that recall practice beat meaning-focused concept mapping at one week. This suggests recall adds something that “deep processing” alone doesn’t explain. See Roediger and Karpicke (2006).
  • Active learning: a large review found active learning beat lecturing in university science, maths and engineering courses (Freeman et al., 2014). Levels of processing offers one plausible explanation, but that review didn’t test why.
  • Re-reading is rated a low-value study method by Dunlosky et al. (2013).
Cite as

Craik, F. I. M., & Lockhart, R. S. (1972). Levels of Processing: A Framework for Memory Research. Journal of Verbal Learning and Verbal Behavior.

DOI: 10.1016/S0022-5371(72)80001-X