Finite Attention Span

I discovered Chris Atherton’s blog and twitter feed this evening. Fascinating, intelligent, relevant writing. Added to blogroll and recommended to all.


The magical number 4

My post on working memory whittled 7±2 down to 5±2; turns out I should’ve settled on 4±1.



Edit: This post was prompted by Olivia Mitchell’s blog entry, which led me to Chris Atherton’s work. Credit where it’s due!

Geoff Petty – Teaching Today (4th Edition, 2009)

Recently, I was pondering the development of questioning technique and was leafing through a copy of Geoff Petty’s ‘Teaching Today’ (a widely recommended reference book for teachers). I was surprised to find that his book has been revised and re-issued as Edition 4 (2009). Unlike a lot of other authors, Petty isn’t bound by his own ideas; he interprets research and accepts improvements. He references Coffield’s Learning Styles report amongst a host of other recent research. I bought a fresh copy of his book immediately. It’s an easy read, filled with golden nuggets – I highly recommended it.

Learning styles – the key points

My article on learning styles was lengthy; the key points were easily lost in the detail and it required effort to work through. Here then, are the key points – the quick, ready reference.

  • Many ‘learning style’ theories have gained popularity and acceptance.
  • UK education professor Frank Coffield headed a team in 2006 responsible for subjecting the most popular of these theories to rigorous academic scrutiny.
  • Coffield’s team found 71 distinct theories; 13 of them were examined in detail. His report largely discredits the concept of learning styles; not one of the most widely (wildly?!) trumpeted theories was considered valid, reliable or recommended for use in today’s classroom.
  • There was however, a positive note: Coffield noted that the development of theories has produced a ‘lexicon’ of terminology – a shared vocabulary which teachers and students can make use of when discussing learning.

Interventions of proven effect

Thinking professionals might wonder how they are best to support learning, given that the (predominantly pseudo-scientific) concept of ‘learning styles’ has been discredited? The answer lies with interventions of proven effect; teachers (and their supervisors) should focus effort on improving and assessing:

  1. Reinforcement
  2. Prior cognitive ability
  3. Instructional quality
  4. Direct instruction
  5. Student disposition to learn

A focus on improvements in these 5 areas produces gains out of all proportion to the completion and analysis of unsupported ‘learning styles’ questionnaires.

4 Questions

If a learner is struggling, then teachers would do well to ask the following 4 questions:

  1. What prior knowledge does the learner have? (exams/recent training/education etc)
  2. How motivated are they? (Not just how motivated do they appear in interview etc. What is their behaviour like? Do words match actions? What evidence supports this?)
  3. How confident are they (going into the learning experience or exam)
  4. Do they demonstrate natural aptitude?

The answers to these questions, duly considered, should really help with the development of individual learning plans. It’s not rocket science, it’s not pseudo-science – it’s proven and recommended; maybe not as funky, but recommended nonetheless.

PS: With reference to targeting improvements in the quality of teaching, often one of the most effective gains to be made is through the enhancement of questioning technique – that’s the subject of my next article!

Learning Styles – a well intentioned, misdirected effort?

Learning styles, we’ve all heard of them, we’ve all filled out a questionnaire or two. Honey & Mumford, Kolb and others are theorists we’ve all heard of. Reflector or pragmatist? Visual/Auditory/Kinaesthetic learner? What about your learners? How do you differentiate for learning styles in your classroom?

Learning styles make us feel good. Discussion opens educational dialogue between teacher/trainer/learner – a lexicon of learning, making us feel professional, as if we’re adding value, promoting the deep-rooted psychological domains of learning. The problem is however, that the learning styles field has been built on foundations of sand. The theories are unproved. Their usefulness and application don’t stand up to academic scrutiny, or peer review.

The whole field of learning styles was comprehensively and professionally scrutinised in 2005/6 by Professor Frank Coffield and his team, funded by the Learning and Skills Research Centre. Their remit was to examine learning styles theories and comment on their usefulness and application in the post-compulsory sector. They “sought to sift the wheat from the chaff among the leading models and inventories … basing conclusions on evidence, reasoned argument and healthy scepticism”

In 16 months of research the team identified 71 models of learning style and categorised 13 of them as major models for detailed examination (matching them against the minimal criteria of internal consistency, test-retest reliability, construct validity and predictive validity). Amongst critiques of 10 other models you can read a detailed review of Honey and Mumford’s Learning Styles Questionnaire (LSQ), Kolb’s Learning Style Inventory (LSI) and the Myers-Briggs Type Indicator (MBTI). It’s difficult to believe that both Honey & Mumford and Kolb’s models (used widely in training/education) were found to meet only ONE of the four minimum criteria and neither received support or recommendation from the review team. Incredible.

Those of us with a penchant for accessible tools often use models that – on face value – appear credible and are well thought of by our peer groups. Prof Coffield’s findings give resounding cause for concern though. His team dismissed some of the most popular models completely, but also made sound recommendations for good practice in the place of learning styles theories. If you’re serious about your practice – indeed, if you’re using some of these models in person – then you really should feel duty bound to read through the work yourself and make an informed choice about whether or not to continue:

In June 2004 the Minister for School Standards, David Miliband, commissioned an independent report to consider the issue of learning. The findings, published by Demos, strongly support Coffield et al.

A bleak picture? Not quite. There’re actually some really useful things we can do to promote learning (and educational dialogue) without resorting to unproved learning styles theories. As Coffield’s team point out, Hattie (1999) detailed effect sizes for classroom interventions, concluding that reinforcement, prior cognitive ability, instructional quality, direct instruction and student disposition to learn (in that order) were by far the most important matters deserving of our attention.

Les Howles’s “Learning Styles: What the Research Says and How to Apply it to Designing E-Learning” is a detail rich summary written in 2006. He indicates a misplaced focus on learning styles and how this focus can be replaced by consideration of “learner attributes that have been tried, tested and proven true (prior knowledge, motivation, aptitudes, and learner confidence related to the content or task to be learned). Interestingly, both “Prior knowledge” and “motivation” sit well within 2 of Hattie’s previously detailed top 5 effect sizes (prior cognitive ability and disposition).

When supporting performance improvements at work I try to consider baseline questions such as:

  1. What prior knowledge does this learner have? (exams/recent training and education etc)
  2. How motivated are they? (Not just how motivated they appear, in interview etc. What is their behaviour like? Do their words match their actions? What evidence supports this?)
  3. How confident are they (going into the learning experience or exam)
  4. Do they demonstrate natural aptitude?

Responses can be matched with efforts towards interventions of best effect – high quality delivery, matched with reinforcement and individual attention.

Learning Styles – a well intentioned, misdirected effort. The research is there and it has addressed us all; we must develop our practice carefully if we are to improve the quality and experience of learning in our classrooms. If you’ve read my article 7±2 you will have noticed that I’ve highlighted two useful ‘chunks’ for your consideration – 5 interventions and 4 questions (both sitting nicely at 5 items ±2). A simple recipe, easily memorised. Simple, but proven.


A popular Urban Legend suggests that 7 digits (±2) is the average amount of information which can be stored for short-term use in working memory.  That’s 7 digits (±2), not a lot at all.

Think of the telephone numbers you know by heart.  Chances are you’ll break those numbers down to 2 or 3 chunks – a chunk of 5 and a chunk of 6 for instance.  Not many people recall telephone numbers as an 11 piece string; we chunk data for immediate access and use sophisticated associations (including visual and auditory cues) for transferring information to long-term memory as required.

7 digits ±2.  Like I said, it’s not a lot.  With words the situation gets tougher.  Evidence suggests that 5 words ±2 is more realistic; a useful rule of thumb being that we can remember (in the short-term) anything we hear which is 2 seconds or less in duration.  2 Seconds?  That’s really not a lot!

There are many useful models used in training and education.  I’ve found a good handful of models really worth remembering for ready use.  In the coming months I’ll detail some of my favoured models along with thoughts on their importance and application.  You’ll find that most of the models contain 5 elements (or less).  I’m a simple man with simple needs; 5±2 is plenty for me.  The process of learning need not be complex at all when we apply credible strategies, using proven models with forethought and a deliberate, passionate approach.