Nick Rose is a ‘Leading Practitioner’ for psychology and research, and teaches at a comprehensive secondary school in Hertfordshire. His blog can be found here and he occasionally tweets as @Nick_J_Rose.
Of the various factors which appear to be driving excessive workload in schools, my suspicion is that the principal among these is uncertainty.
For a school leader, it is the uncertainty of whether an Ofsted inspection team will agree with their judgements about the quality of teaching within a school. The stakes for getting this wrong are very high for the school leader but also for the future of that school. Recruitment of teachers can be damaged by a poor judgement – with ‘requires improvement’ schools ‘locked out’ of teacher training and having second pick of new teachers entering the profession. This can trap schools into a vicious cycle, where their status as an RI school makes it more difficult for them to recruit the best teachers (and thus harder to move out of the RI category). It also seems to affect student recruitment to the school with middle-class parents more likely to refer to Ofsted reports and attainment data when selecting a school. This may mean that an RI rating leads to fewer middle-class parents (who appear more likely to support their children through tutoring, for instance) sending their child to that school. Given that Ofsted grading correlates strongly attainment of children prior to arriving at the school, this also may make it a harder journey for that school to lift itself out of a category. Ofsted appears to have recently acknowledged this issue, admitting that recruiting pupils with higher ability makes it easier for a school to achieve a good or outstanding judgement.
We have no data for the reliability of Ofsted judgements, but a recent analysis by the Education DataLab might suggest there are significant problems. They asked, “Is an Ofsted judgement a lagging or leading indicator of school performance?”. They found that where the quality of teaching would appear be improved in a school compared to its past performance, Ofsted did not appear to spot this. Likewise, an Ofsted judgement did not appear to predict an imminent deterioration in performance in a previously well-performing school. If past performance (rather than current capability) strongly influences Ofsted judgements, what does this mean for the school improvement efforts of schools in a category? Given that Ofsted may or may not spot a genuine improvement in teaching, how does a school work towards an improved judgement? This inherent uncertainty acts to drive up workload (as school leaders have to provide more and more ‘evidence’ they are doing their job to protect themselves against the vagaries of judgement) and ultimately cause the reported problems in the recruitment of school leaders.
These high-stakes for schools lead to high-stakes appraisal system for teachers. The problem for a teacher is the inherent uncertainty as to whether evaluations of the quality of their teaching will lead to questions about their professional capability. The reliability of typical measures of teacher quality is poor. For example, looking at correlations between teacher evaluation and long-term successful outcomes for students within mathematics, observations were 0.24, student surveys 0.36 and even value-added scores were only 0.69. It would appear the value added associated with exam grades correlates most strongly with teacher effectiveness, but this relationship is surely weakened by the fact that that prior attainment also appears to influence progress measures and there are questions about the accuracy of marking from the fact that 90,000 exam grades were changed this year. The frequent use of poorly correlated measures like observations and student surveys generates even more uncertainty. This acts to drive up workload (as teachers have to provide more and more ‘evidence’ they are doing their job to protect themselves against the vagaries of judgement) and ultimately cause the high levels of stress reported by teachers.
The use of any sort of statistical data involves an implicit uncertainty which is rarely reflected in Ofsted or teacher quality judgements. At a fundamental level, any decision is open to two types of error: Type 1 errors (false positives) where an underachieving school or teacher is rated as good or highly effective and type 2 errors (false negatives) where a good school or an effective teacher is rated as underachieving. Minimising one type of error tends to increase the other type of error; so systems with ‘tougher’ standards will likely unfairly condemn more schools or teachers as ineffective. Until schools and school evaluation systems start to acknowledge the fundamental uncertainty inherent in such judgements and continue to attach such high-stakes to them, this uncertainty will continue to drive up workload in schools.