How should we evaluate CPD?


Bridget Clay argues that pupil need, not teacher practice, should be at the heart of CPD.

Only six per cent of primary and eight per cent of secondary schools claim not to evaluate CPD at all according to an NFER Teacher Voice study in 2009. 

Many a twilight or INSET session is followed by a staff survey of varying length and detail. However, given that these often appear a very short time after the session, what exactly is being evaluated? 

Teacher experience of a training session is not what drives school improvement. Surely the essence of CPD is how we can develop staff to enable the best possible outcomes for our students? Yet only seven per cent of all schools evaluate CPD on pupil progress according to the same NFER survey.

There are two issues to be pulled apart here. First, the great majority of CPD is focused around teacher practice. Common examples include CPD focused on differentiation or Assessment for Learning. It is not uncommon for these to be repeated year after year. However, addressing teacher practice does not necessarily have a significant impact on pupil outcomes and determining CPD priorities by teacher practice, as opposed to pupil need, is skewed. 

Pupils should be at the heart of a school’s focus, CPD included. That is not to say that teacher experience is unimportant; CPD should improve the confidence and skill of practitioners. However, teachers choose the profession in order to improve pupil learning and outcomes. CPD that is focused on improving pupil learning can only benefit both teachers and schools.

Second, as a result of the emphasis on practice rather than learning, CPD is very poorly evaluated, as it is hard to quantify, relies heavily on observations and does not directly relate to school improvement. CPD needs to be turned on its head; it should be driven by the needs of the pupils, and should be focused around how to meet those needs. Evaluation should therefore also be based around pupil learning.

It is widely acknowledged that this is a challenge. One of the most difficult aspects to overcome is how to evaluate which specific teaching approach has had an impact on learning. This is why professional development should be focused on a very specific cohort of pupils and a very specific need.

In the National Teacher Enquiry Network (NTEN), practitioners are supported to implement Lesson Study, a model of CPD proven to be very effective. This involves identifying a specific group of pupils and a particular need they have, carrying out an evidence-informed change, and measuring the impact carefully on three particular case pupils. 

Observations are learner, rather than teacher-focused, and all evaluation is based around changes in learning. We suggest that all CPD should be done and evaluated in this way.

Effective CPD evaluation should involve a variety of methods. Standardised national tests, surveys and interviews are effective, while learner-focused observations allow for specific measures to be recorded, such as the number of times that an activity occurs.

Comparison or control groups are particularly useful for evaluating impact. Ethically, it is important that these are only used when you are not entirely certain that the intervention idea is superior to existing practice, although it is equally important not to merely assume that interventions are superior without testing. In previous studies, ideas that have previously been thought to be superior turned out not to be.

Evaluating CPD is not without its challenges. Yet only by focusing on learners and evaluating as rigorously as possible can practitioners truly inform and develop their practice.

  • Bridget Clay is the National Teacher Enquiry Network Support Officer at the Teacher Development Trust. She is a former maths teacher.

Further information
The Teacher Development Trust is the national charity for effective professional development. Find out more at or access the TDT’s free database of CPD at For details of the recently launched NTEN, visit



Please view our Terms and Conditions before leaving a comment.

Change the CAPTCHA codeSpeak the CAPTCHA code
Claim Free Subscription