Saturday, 2 March 2013

Testing teachers

I wrote this blog post for Independent Voices on why improved test scores are a far better measure of success than student surveys

Good teaching is at the heart of good schools. We have done a lot to improve the quality of new teachers, but there has been much less focus on the quality of the existing workforce. Yet, while 35,000 new teachers enter the profession each year, the teacher workforce is 440,000-strong.

Schools need to make the most of teachers’ talents if young people are to get a decent education. For a disadvantaged pupil, an excellent teacher can deliver the equivalent of 1.5 years learning in a year, whereas a poor teacher contributes just half a year: the difference is a whole year of a child’s education.

That’s why it is important we evaluate the contribution that teachers are making and can make with the right support. A new Sutton Trust study, Testing Teachers, shows that the contribution that teachers make to improving exam and test results is the most reliable way to predict a teacher’s long-term success.

The study, by Richard Murphy of the London School of Economics, drawing on the latest international research, shows that improved test scores are nearly twice as effective as student surveys and nearly three times more effective as classroom observations.

But schools can’t simply look at a single year’s test scores to assess performance. A reliable and fair approach requires a sensible combination of these and other measures taken over several years, and might also include teachers’ contributions to sports and school trips.

When Labour introduced performance related pay in 1999, it did so within a very bureaucratic framework that didn’t work as intended in most schools. By contrast, the education secretary Michael Gove is hoping that leaving schools to develop their own systems will improve results and see the best teachers more effectively rewarded.

But without the right systems in place, schools may be no readier to do so now than they were in the past. So what are the characteristics of an effective system of teacher appraisal?

Most importantly, it should involve clear standards, fairly and consistently applied. External advice can be helpful in getting this right, and could assure staff of its fairness and governors of its robustness.

Teachers or school leaders involved in evaluation should be properly trained, and should discuss their evaluation fully with the teachers concerned.

When using exam or test results, it is important to focus on value added rather than absolute results, as they are the most objective and comparable assessment of a teacher’s contribution. It is also important that the baseline for such comparisons is sufficiently robust.

With classroom observations – where teachers or school leaders witness teaching in practice – the report suggests that those designed to help a teacher improve should be carried out separately from those used for appraisal, as this is more likely to promote honest feedback.

Pupil surveys can also be used – particularly with older pupils – as they are the ones in most day-to-day contact with teachers, but when they are they should be clearly structured, be age appropriate, and should complement other measures.

Getting all this right can have real benefits for pupils and teachers alike. Earlier research for the Sutton Trust has shown that if we were to raise the performance of the poorest performing tenth of teachers to the average, we would move into the top rank of the OECD’s PISA tables internationally.

But there is a more compelling reason: by improving the quality of our teachers collectively, we can ensure that every child has a decent education, and is not held back by poor teaching. That is a goal well worth pursuing.
 

No comments: