Independent schools in state v private research were 'defined' by the group commissioning the report

Janet Downs's picture
 8

Two months ago the media was full of research by Durham University which suggested attending a private school equalled two years extra education at age 16 or 0.64 of a GCSE grade.  That conclusion was seized upon by the Independent Schools Council (ISC) who had commissioned the report to claim the superiority of private schools.

But the ISC and media coverage missed researchers’ warnings.  The results weren’t as clear cut as headlines claimed.  The researchers studied existing literature which considered the existence of any private school advantage.  The conclusions in the literature were inconclusive: the  'jury is still out' on whether attending a private school enhanced academic achievement.   And our Henry Stewart cast doubt on the assertion that 0.64 of a GCSE grade equalled two extra years schooling.

I wanted to know whether the Durham researchers looked at GCSE results from ALL independent schools or only those who were members of groups such as the Headmasters' and Headmistresses' Conference (HMC) for which ISC acts as an umbrella.     Membership of these groups is not open to all fee-charging schools but restricted to those of a particular 'quality'.  If the research were limited solely to schools in these affiliated groups, then this would surely affect the results.   There are over 1000 independent schools which are not members of such organisations.   If these non-affiliated schools were included then it would likely reduce the alleged independent schools advantage.

The report itself gave a clue.  It said: ‘The set of independent schools involved in the study were defined by the ISC. They provided a list of independent schools.’

This confirms that the independent schools used for comparison were chosen by the ISC, the report’s commissioners. 

I wanted to find out if this list contained non-affiliated schools and, if so, how many.  So I sent a Freedom of Information request to Durham University.  This asked for the following:

1 How many affiliated independent schools were surveyed. 
2 How many non-affiliated independent schools were surveyed. 
3 The proportion of surveyed independent schools which selected their intake for ability.

The University replied saying it did not have the information I requested. 

Since Durham University claims not to have this data then I can only presume the independent schools were ones affiliated to the ISC and, being mostly selective, were likely to have GCSE results higher than state schools which cater for the whole ability range. 

When the report was published, Julie Robinson, General Secretary of ISC, acknowledged it was difficult to compare fee-charging and state-funded schools and there was ‘much excellence to be seen in school of all types’.  But she added that the ‘ground-breaking report from such a world-renowned and respected research unit at Durham University really does give us solid ground to say that based on academic results, independent schools are worth paying for.’

But if the independent schools used for comparison were mostly those who select for ability and did not include non-affiliated schools, then how can Durham’s research be described as ‘solid evidence’ that schools in the independent sector are ‘worth paying for’?

UPDATE 17 May 2016.   We were right to be sceptical about this ISC commission report.  Se analysis by Nick Hassey, Teach First, on the BERA blog.

The above article has been amended to make it clear the conclusion that the jury was still out about any private school advantage was from the literature which was studied by the researchers.

 

Share on Twitter Share on Facebook

Be notified by email of each new post.





Comments

Jane Eades's picture
Mon, 18/04/2016 - 14:32

I have also heard stories of independent schools simply refusing to accept the cheques for pupils in the year running up to GCSE if they are unlikely to achieve results which will boost the school's reputation.  The other dodge is using IGCSE which is not reported in the DfE tables.  So, all in all, this sounds like a rather dodgy piece of research.


ian thompson's picture
Mon, 18/04/2016 - 16:04

In the interests of clarity it should be pointed out that the Durham researchers said that the jury was still out in their summary of the previous UK evidence and before describing their results and conclusions:

'It is clear that studies carried out in the UK tend to support the view that independent schools confer higher academic ability on students than state schools but some of the studies had clear limitations. For example, some of them did not control for students’ prior ability and none of the studies looked at differences between the two sectors from early stages of students’ education... Therefore the jury is still out with regard to the true extent to which attendance at independent schools in England enhance the academic achievements of students when student and school-level differences are accounted for'

Having addressed the impact of prior ability themselves their conclusion was that:

'Differences existed between the academic achievement of independent and state schools at every age group even after underlying differences between the two sectors had been controlled for'

As data was analysed for four year olds, eight year olds, ten year olds and sixteen year olds, GCSE results are only one part of the picture. However the tables on page 16 show that their samples were representative of the state and independent school populations as a whole, underlying the credibility of their GCSE data.


Janet Downs's picture
Tue, 19/04/2016 - 08:35

Hi, Ian - you're right to point out 'the jury is still out' re any academic advantage supposedly offered by attending an independent school.  That important point was missed by both the ISC and the media which churned the ISC press release.   

The ISC press release concentrated entirely on results at 16 so I confined my comments to that and not an analysis for 4, 8 and 10 year olds.  In any case, the same problem arises when comparing results from Performance Indicators in Primary Schools tests (PIPs) between independent and state schools at such an early age.  Children entering reception in private schools are hardly likely to encompass the whole ability range.   PIPS tests are highly regarded but I don't think they're used in all schools in the same way the Early Years Foundation State Profile was.  

Durham, which developed PIPS, said many items in PIPs were 'language based' and therefore 'It follows that further analysis of PIPS data to compare EAL children with others at the start of the year is potentially problematic.'  (Page 21 here).   I would add children whose language development is poor to this group.  It follows, then, that comparing schools which are less likely to admit pupils whose language development is poor (private schools) with schools which have to admit all pupils irrespective of their communication skills (state schools) is similarly problematic.

And there's the nub of the problem.  The names of the independent schools used in the survey were submitted by ISC.  It's not likely that this list would have contained non-affiliated independent schools.  The research, then, can't be said to cover the entire independent sector yet the press release and subsequent reporting implied attending any independent school would confer an advantage.

Update 08.41:  This has been amended to correct typos and garbled grammar.

 


Janet Downs's picture
Tue, 19/04/2016 - 08:39

Ian - there's a further problem.  Durham  writes on the page you cited, 'The population of independent schools consisted of all independent schools that had GCSE results in the data acquired from the NPD [National Pupil Database]'.  But the NPD only list data re pupils in independent schools 'where available'.   The NPD data, then, can't be said to represent all independent schools since results for many of them (number unknown) would not be in the NPD data.


Jane Eades's picture
Mon, 18/04/2016 - 16:18

There seems to be some confusion between academic achievement and academic ability.  I remember some research years ago which demonstrated that, for similar A Level results, students from states schools performed much better at University and achieved better degrees.


Janet Downs's picture
Tue, 19/04/2016 - 09:12

Janee - you're right.  A summary of the research which found state-educated pupils outperformed their equally-qualified peers from private schools at uni is here.   At the 'elite' level, educational background, private or state, made no difference to the level of degrees awarded neither Oxford while nalysis by Cambridge Assessment found 'independent school-leavers at Russell Group institutions are a third less likely to get a good degree than state school students with similar A-level results'.


ian thompson's picture
Mon, 18/04/2016 - 17:37

Smith and Naylor showed in 2001 that pupils who had been at independent schools were 6.9 and 8.6% (females and males respectively) less likely to obtain a good degree than state educated pupils with the same A level grades, but the effect weakened the higher the grades. Female gender and higher social class were positively associated with increased probabilities of obtaining a good degree of a similarly small magnitude. These findings have been replicated since.

Considering ability curves and achievement puts this into perspective because state and privately educated A level candidates don't achieve the same A level grades and derive from entirely different ability spectra for their education sectors. The privately educated cohort represents almost the whole ability curve encompassing strong and academically weak pupils and achieves BBB on average at A level. State educated pupils from the top half of the ability curve achieve on average CCC. Looked at this way, comparing pupils from the two sectors with the same grades leads to a remarkably small difference in subsequent performance.

http://www2.warwick.ac.uk/fac/soc/economics/staff/ranaylor/publications/...


Nigel Ford's picture
Tue, 19/04/2016 - 20:37

I wonder how close to home Durham University conducted their research.

On A'level criteria, the local comp, Durham Johnston, outperforms the city's public school of the same name, by nearly 15 points per A'level entry, despite there being a parity for 6th form entry criteria into the 2 schools in question.


Add new comment

Already a member? Click here to log in before you comment. Or register with us.