School improvement

There is now a widely held view that the most effective strategy for improving countries’ educational performances is to improve the day-to-day work of schools. This view follows several decades of significantly increased expenditure on school education, including the funding of major reforms and targeted intervention programs, often with little or no accompanying evidence of improvements in the quality or equity of educational outcomes.

The ultimate goal of school improvement is to improve outcomes for students.

Promoting school improvement

In an effort to ‘drive’ improvements in the day-to-day work of schools, many education systems have introduced rewards and/or sanctions tied to school results. Rewards sometimes take the form of financial incentives; sanctions may include increased system intervention, the replacement of the school principal or, in extreme cases, school closure. The theory of action underpinning these schemes, which tend to follow models adopted from the world of business, is that when rewards or sanctions are attached to desired results, greater employee effort ensues, resulting in improved outcomes.

An example of a reward scheme of this kind is the Australian Government’s Reward for School Improvement initiative, introduced as an election commitment in 2010. Under this initiative, $275.6 million in reward payments will be provided between 2015 and 2020 to Australian schools that can demonstrate improved student results.

Despite their widespread use, results-based incentive schemes have a disappointing track record. In businesses, results-based incentives appear to be effective only in motivating relatively low-level work. A recent US review of test-based incentive schemes in schools concluded that the benefits have been quite small and highlighted the many unintended ways in which financial rewards distort the work of schools.

Part of the reason for the limited success of results-based incentives is that they ignore the research on human motivation. For example, there is evidence from psychology that paying people for things they would have done anyway can lower performance. But perhaps more importantly, results-based incentives are based on the assumption that employees know how to improve and that what is lacking is effort. In education and in business it is now recognised that this assumption undervalues the importance of capacity building and organisational culture. As Richard Elmore observes, ‘people in schools already are working pretty reliably at the limit of their existing knowledge and skill and giving them information about the consequences of their practice will, in general, not significantly improve that practice’.

Maintaining a focus on outcomes

Nevertheless, the ultimate goal of school improvement is to improve outcomes for students. Improvements in school practices and processes are largely meaningless if they do not lead to better outcomes for the students who attend them. For this reason, reliable outcome measures will continue to be essential to evaluations of school improvement.

Experience in school systems that have introduced results-based incentives is that when narrow measures of student outcomes – for example, only literacy and numeracy test results – are used to make high-stakes decisions, the behaviours of teachers and schools change accordingly. These changes are sometimes positive and intended, but they also include a variety of ‘gaming’ behaviours designed only to increase test results. Schools sometimes assign their best teachers to the year levels in which tests occur, inappropriately narrow the focus of teaching, withhold less able students from testing and, in extreme cases, engage in cheating practices such as ensuring that relevant information is displayed on classroom walls. The consequence is ‘score inflation’, with results on high-stakes tests not being matched by results on low-stakes tests of the same general content.

A response to these observations has been to collect data on a broader range of outcome measures, including school attendance, school completion, disciplinary actions, Year 12 results and post-school destinations.

Importantly, measures of school improvement require evidence of change over time. This is because measures of student achievement at a point in time are strongly influenced by students’ backgrounds and starting points. Improvements in teaching and learning can be inferred from changes in student performance (for example, changes over time in Year 5 reading levels in a school). But even here, caution is required. Cohorts of students in a school can vary from one year to the next, meaning that school improvements are best inferred from long-term trends in achievement levels rather than from year-to-year fluctuations.

Measures of improving school practices?

School improvement requires changes in the work of a school – particularly enhancements that lead to better outcomes for students. For this reason, an argument can be made for defining school improvement not only in terms of improving student outcomes, but also in terms of improving teaching, learning and leadership practices.

Strong foundations exist for the development of practice-based measures of school improvement.

At the present time, there are no agreed practice-based measures of school improvement that could be used alongside outcome-based measures of improvement. However, strong foundations exist for the development of practice-based measures. There is a high level of consensus in the research literature on the general characteristics of highly effective schools, classroom teaching and school leadership. Most school systems have incorporated this knowledge into their school improvement frameworks and regular school review processes.

If rewards are to be provided to schools for evidence of improvement, then they should be tied to matters over which schools have direct control; they should make transparent the relationship between rewards and the day-to-day work of schools; and they should promote highly effective, evidencebased practices. ACER’s discussion paper for the Australian Government proposes a set of principles to underpin the Reward for School Improvement initiative. Specifically, rewards need to:

encourage schools to evaluate and monitor their own ongoing improvement;
be based in part on evidence of improved student outcomes;
be based in part on evidence of improved school practices;
build on and enhance existing systemic school improvement efforts; and
be based on fair comparisons.

ACER has developed a Teaching and Learning School Improvement Framework that addresses these points. Read the article 'A framework for school improvement' for further information.

Read the full report:
ACER's discussion paper for the Commonwealth Department of Education, Employment and Workplace Relations, Measuring and Rewarding School Improvement, by Professor Geoff Masters, is available at <research.acer.edu.au/policy_analysis_misc/15/>

School improvement