Harry Fletcher-Wood, Associate Dean at the Institute for Teaching

It seems there’s a consensus when it comes to what effective professional development looks like.  Reviewers, like Laura Desimone and Linda Darling-Hammond, argue in favour of this consensus when considering what should be included in the design of professional development for teaching, on the grounds that it’s backed by strong evidence. It includes priorities like focusing on student outcomes, offering sustained development and prioritising collaboration.  However, there’s a problem: the theory isn’t watertight, and the practical guidance it offers isn’t always effective.

Why the consensus isn’t always effective?

Most professional development programmes include the whole consensus recipe: they are collaborative, sustained, subject-specific, and so on.  But which of these aspects really makes a difference?

The consensus is particularly problematic for a teacher educator.  It’s rarely sufficiently specific to be useful: if ‘collaboration’ means any time teachers are working together, what kind of collaboration works best?  

If – for example – sustained, individualised coaching is the most effective choice, but we can’t afford it, should we try short coaching, coach fewer people, or give coaching in bigger groups?  What would happen if we moved our coaching online?  The consensus view can’t answer the questions that matter most to teacher educators.

Lots of studies seem to work when the designers implement them in a few choice schools, but the effects diminish or disappear when the programme tries to grow.

What happens when it’s applied in schools?

When researchers apply the consensus in schools, the results are unpredictable.  Many trials have struggled to show professional development works and have suffered implementation difficulties.  One year-long maths professional development programme offered summer training, collaborative groups and video coaching: teachers’ behaviour changed, but student achievement didn’t.

Another trial lasted three years. This focused on problem solving and the examination of student work: teachers liked it and reported learning more maths content themselves, but their teaching didn’t change and nor did their students’ results.

The opposite appears to be true too. Other programmes which haven’t included all of the consensus ingredients have proved successful.  Individual coaching has had an effect on teachers in a matter of weeks, and interaction coaching has proved extremely powerful despite including no subject-knowledge element.

If programmes which follow the consensus view of good professional development sometimes don’t work, and programmes which don’t follow the consensus view sometimes do, surely we have to question whether the consensus really offers teacher educators the guidance they need?

We are enthusiastic about professional development, but we have to recognise that it tends not to have the impact we want.  Has following the consensus view created guidelines of doubtful usefulness, and led us to focus on the design features, rather than our goals?  Do we have the guidance we need to be equipped to offer effective professional development?  

If you’re interested in reading about this subject in a bit more detail have a look at the first section of my paper, Designing Professional Development for Teachers, What Is the Consensus View? 

In our next blog….

We suggest some alternative ways of thinking about professional development design.

Find out about how we’re putting our research on professional development into action on our Fellowship in Teacher Education, run by Harry.