Of course, you should not take everything people say to you at face value. Heads are supposed to know what’s going on - it’s part of the job. None the less, if inspections were just about school improvement, then their assertions would be profoundly worrying.

But for well over a century, inspection’s main purpose has been to serve as a system of public accountability. “Do good as you go” was a motto of many HMIs, but school improvement was not their main task. As a former chief inspector has remarked, the concern to put it on the inspectors’ agenda is a phenomenon of the Nineties. Not surprisingly, many questions about exactly how one harnesses inspection to the cause of improvement have still to be answered.

When the first Office for Standards in Education cycle has been completed, some half a billion pounds will have been spent. Have the gains in knowledge and understanding really been so insubstantial that we could have achieved the same results by cheaper and less stressful means? Whatever the reality, the fact that so many heads think this way has to be taken seriously.

Preparing for OFSTED has become a national obsession. Some schools start more than a year ahead, and all report on the intensity of the week of the “performance”. Then, exhausted, they rake over the apparently predictable reviews from the predictable inspectors, taking some satisfaction from the things that were missed as well as those that were noted with approval.

There is nothing wrong with a school seeking inspectors’ accolades. What is depressing is that, at a time when it is widely recognised that the whole system needs to move up a gear, heads have been encouraged to seek a “good” report when what most really need is a constructive analysis.

A few schools did learn something from inspection. They were surprised by what the inspectors told them - shocked might be a better word. Schools in trouble do seem to have had greater difficulty in taking stock. For them a constructive analysis (preferably offered in a manner which enables them to retain a modicum of dignity) is essential. Unfortunately, what many thought they got was just a “bad” report.

The challenge for the next OFSTED cycle is to build a system combining “light touch” accountability with maximum support for change. Where might one start?

First, with the notion that an inspection is something that is usually conducted between Monday morning and Thursday night, with the main conclusions being presented on Friday. Schools are complex organisations and take varying amounts of time to understand. Something more flexible than what we have now is required. Inspectors might start by staggering visits over a longer period.

The need for accountability and the desire to aid improvement may be in tension here. For the purposes of accountability we simply need to know whether a school has serious weaknesses. We are deluding ourselves about the precision of school evaluation if we attempt to make assessments too finely-grained.

The present comprehensiveness of inspection is only really necessary for those schools which have serious weaknesses. The vast majority (perhaps 90 per cent) do not fall into this category and for them it would be better if inspectors concentrated on unlocking just two or three key issues for development. Hence in any future OFSTED cycle it would be foolish to tell schools again what they already know.

Second, the number of criteria on which inspectors are required to collect evidence should be cut back to a very modest core. The previous “ring binder” version of the inspection manual was falling apart under the strain of successive directives which usually complicated and only rarely clarified. Even the recently-revised manual needs drastic pruning. There is a simple trade-off between the number of questions inspectors ask and the quality of the insights they obtain. The broader the range of “evidence” they attempt to collect under time-constrained conditions, the more superficial they become. In the process of rewriting the manual yet again, OFSTED might also want to explore some of the fundamental methodological issues on which inspection is perched - most pressingly, the reliability of inspectors’ judgments.

Third, we need a different kind of report that is more illuminating to read and which schools can learn more from. Schools will accept the “hard” messages, if they think inspectors have put in sufficient effort to find out what’s really happening. However, they also want to have their eyes opened to the alternatives - there is rarely one right way to school improvement.

Three years ago, readers of The TES were promised that the “skilled observation” that inspectors undertook, would “throw into particularly sharp relief the highs and lows of pupils’ school life - the magical, inspired or simply very competent lessons they experienced as well as any moments of torpor, incomprehension or alienation”. This was a worthy aspiration but not, unfortunately, one which survived. We ourselves have yet to read any accounts of magical or inspired lessons in inspectors’ outpourings. Yet the sentiment was undoubtedly right; teachers are motivated by visions, not mandates.

Fourth, inspectors need to report on schools’ capacity to improve. Heads frequently complained to us that the inspectors had ignored or played down how much had already been achieved.

Fifth, the qualities of our inspectors should be examined in terms of their fitness for purpose. Do they have substantial experience of the problems facing the kinds of schools they are inspecting? Do they have first-rate analytical skills? And crucially, can they inspire confidence in their findings and motivate teachers to accept their challenges? In the rush to recruit enough inspectors to provide every school with its educational MoT, too little attention has been paid to these concerns.

Sixth, inspectors must rebuild the partnership with the profession. A limited element of self-evaluation may help but this is not, in our view, the main way ahead. In too many schools it is still seen as another hoop to jump through and a further distraction from the real business of teaching. OFSTED needs to restore a greater element of trust to the whole inspection process.

The following changes would be easy to implement and might help.

* Someone who the school sees as a “critical friend” should be made a member of the inspection team along the lines the Further Education Funding Council has successfully piloted in post-16 institutions.

* Schools should be given the right of reply by having comments they want to make on their reports added as an appendix. If the inspectors’ findings are to service school improvement, then they must be robust enough to survive criticism.

* Schools should be asked to identify a member of their inspection team to act in a post-inspection support role. Separating inspection from advice has been taken too literally. Those who generate insights from inspection should play some part in the next stages of a school’s development.

The concern to raise standards provides a long-term agenda. History will judge OFSTED by its legacy of school improvement, not by today’s performance.

John Gray is director of research at Homerton College, Cambridge, and Brian Wilcox is a professorial research fellow at Sheffield University’s Division of Education. A book based on their Economic and Social Research Council-funded study, Inspecting Schools: Holding Schools to Account and Helping Schools to Improve, is published this week by the Open University Press.