Researchers have questioned whether a study which identified the only “truly effective” type of headteacher can be applied across the school system.
The study, published in an article entitled “The one type of leader who can turn around a failing school”, has been widely cited since it appeared on the Harvard Business Review (HBR) website in October.
The research categorised headteachers into five groups such as “surgeons”, whose schools enjoy dramatic short-term improvement at the cost of longer-term setbacks; and “architects”, who “quietly redesign the school and transform the community it serves”. “Architects” were said to be the only genuinely effective type of leader.
The high-profile study was based on data taken from 160 secondary academies, both before and after the tenures of 411 headteachers, over periods of between seven and nine years from 2007.
‘Architects’ were said to be the only genuinely effective type of leader
It claimed to identify a clear link between the type of leader and their subject area. For example, the study said that 85 per cent of “surgeons” had a PE or religious studies background, and 68 per cent of “architects” had studied history or economics at university and had normally worked in industry for 10 to 15 years before becoming teachers.
However, new research by Education Datalab, seen by TES, questions whether the findings can be applied across a wider range of schools, and should guide how they recruit leaders.
Academics at Education Datalab used the government’s school workforce census to establish whether, across all secondaries in England, and specifically those rated less than “good” by Ofsted, there was a link between a head’s subject and their school’s performance.
‘Unusual sample’
In what Education Datalab admitted was a “quick and dirty” analysis, its academics said: “Subject background of senior leaders is not an important factor in explaining variation in school performance.”
Similarly, they found that while headteachers in certain subject groups tended to have spent more time outside teaching before qualifying, they had spent less time than those in the HBR article, and were therefore not necessarily “architects”.
Becky Allen, of Education Datalab, said: “The HBR article clearly recruited a quite unusual sample of 160 schools and I have little doubt that the findings hold for those schools over that time period.
“However, a research study is only useful to schools and policymakers if the findings are applicable beyond the sample of schools on which the original research was conducted. We cannot replicate the findings in a wider sample of schools, which makes us uncertain that they are more generally applicable.
We cannot replicate the findings in a wider sample of schools
“We hope others will now come forward to try to replicate the original findings.”
Liz Mellon, one of the HBR article’s authors, said the team started by looking at schools’ results, and then the actions of their headteachers, which formed five clusters, and then looked at their backgrounds.
“Education Datalab are starting at our ending. If they followed our methodology, and had access to equivalent data, their findings might be different,” she said.
Dr Mellon said that her group would publish their methodology “soon”, and added that they hoped attention would then turn to whether changes need to be made to the way that headteachers are found, recruited and appointed.
“Whether our findings will be found to be applicable outside academy schools remains to be seen, but attempted refutation of our findings should not be based on a ‘quick and dirty analysis’,” she said.
