A new analysis has found a headteacher’s performance is not linked to their subject background – challenging previous research that claimed heads of failing academies can be categorised into five different types.
Researchers from the Centre for High Performance published a report last year that claimed “architects”, often a former economist, historian or music teacher, were the most effective headteachers in “failing” schools.
The study, published in the Harvard Business Review, found “surgeons”, likely to have studied PE or RE, caused rapid turnaround but long-term damage. Other heads were categorised by performance type as “soldiers”, “philosophers” or “accountants”.
While some experts claimed the study, covered by Schools Week, was “groundbreaking”, academics have questioned the findings – including raising concerns over the methodology not being released.
A new analysis by Education Datalab has now further challenged the study, revealing that subject background of senior leaders is not an important factor in explaining variation in a school’s performance.
The data experts also found that difference by pay according to headteacher type did not match up to the Centre for High Performance’s findings.
Dr Rebecca Allen, director of Education Datalab (pictured above), said it was important that other researchers interrogate the original findings which “if generally true, would re-shape the advice we should give governing bodies about how to recruit school leader”.
Not publishing the methodology is a major shortcoming”
But she said this wasn’t straightforward because a methodology for the study has not been published.
Allen told Schools Week the “major shortcoming” of the initial reports was “simply that they choose not to publish any methodology anywhere, meaning that nobody is able to access whether their analysis is robust or valid”.
She said the Datalab study could therefore not replicate the findings, but Allen crunched data from the school workforce census, filtering out schools rated “good” or “outstanding”.
She then looked at the annual change in GCSE results for every senior leader or headteacher from 2010 until when they left the school, or until 2015 if they did not leave their post.
On performance during the headteacher’s tenure at a school, Datalab found subject background was not an important factor in explaining variation in school performance.
The blog added: “There is certainly no evidence that leaders with a PE, RS or biology background have good GCSE improvements.”
Datalab also found there was no difference in GCSE grades over the long-term after a headteacher had left a school, based on their subject background either.
The Centre for High Performance study found the “architects”, the most-effective head type, were paid the least of any group. Meanwhile the most damaging “surgeons” were likely to be paid the highest salaries.
But Allen found, while there are some differences in pay by subject, the analysis could not match the original study’s findings.
Finally, Education Datalab challenged the claim by researchers Ben Laker and Alex Hill that time spent outside education before becoming a school leader improved GCSE results.
“The correlation between years before qualifying and GCSE improvement (either during or after tenure) is essentially zero.”
The original research found English teachers were visionary “philosophers” who had little impact on school attainment and made up the majority of all headteachers.
The “accountants” were former maths teachers who encouraged financial viability but were ineffective for pupil attainment. And the “soldier”, usually chemistry or IT teachers, cut costs but demoralised staff.
Both Laker and Hill did not respond to Schools Week’s request to comment.