High Attainment in the 2014 Secondary and 16-18 Performance Tables

.

This is my annual analysis of high attainment and high attainers’ performance in the Secondary School and College Performance Tables

Data Overload courtesy of opensourceway

Data Overload courtesy of opensourceway

It draws on the 2014 Secondary and 16-18 Tables, as well as three statistical releases published alongside them:

It also reports trends since 2012 and 2013, while acknowledging the comparability issues at secondary level this year.

This is a companion piece to previous posts on:

The post opens with the headlines from the subsequent analysis. These are followed by a discussion of definitions and comparability issues.

Two substantive sections deal respectively with secondary and post-16 measures. The post-16 analysis focuses exclusively on A level results. There is a brief postscript on the performance of disadvantaged high attainers.

As ever I apologise in advance for any transcription errors and invite readers to notify me of any they spot, so that I can make the necessary corrections.

.

Headlines

At KS4:

  • High attainers constitute 32.4% of the cohort attending state-funded schools, but this masks some variation by school type. The percentage attending converter academies (38.4%) has fallen by nine percentage points since 2011 but remains almost double the percentage attending sponsored academies (21.2%).
  • Female high attainers (33.7%) continue to outnumber males (32.1%). The percentage of high-attaining males has fallen very slightly since 2013 while the proportion of high-attaining females has slightly increased.
  • 88.8% of the GCSE cohort attending selective schools are high attainers, virtually unchanged from 2013. The percentages in comprehensive schools (30.9%) and modern schools (21.0%) are also little changed.
  • These figures mask significant variation between schools. Ten grammar schools have a GCSE cohort consisting entirely of high attainers but, at the other extreme, one has only 52%.
  • Some comprehensive schools have more high attainers than some grammars: the highest percentage recorded in 2014 by a comprehensive is 86%. Modern schools are also extremely variable, with high attainer populations ranging from 4% to 45%. Schools with small populations of high attainers report very different success rates for them on the headline measures.
  • The fact that 11.2% of the selective school cohort are middle attainers reminds us that 11+ selection is not based on prior attainment. Middle attainers in selective schools perform significantly better than those in comprehensive schools, but worse than high attainers in comprehensives.
  • 92.8% of high attainers in state-funded schools achieved 5 or more GCSEs at grades A*-C (or equivalent) including GCSEs in English and maths. While the success rate for all learners is down by four percentage points compared with 2013, the decline is less pronounced for high attainers (1.9 points).
  • In 340 schools 100% of high attainers achieved this measure, down from 530 in 2013. Fifty-seven schools record 67% or less compared with only 14 in 2013. Four of the 57 had a better success rate for middle attainers than for high attainers.
  • 93.8% of high attainers in state-funded schools achieved GCSE grades A*-C in English and maths. The success rate for high attainers has fallen less than the rate for the cohort as a whole (1.3 points against 2.4 points). Some 470 schools achieved 100% success amongst their high attainers on this measure, down 140 compared with 2013. Thirty-eight schools were at 67% or lower compared with only 12 in 2013. Five of these boast a higher success rate for their middle attainers than their high attainers (and four are the same that do so on the 5+ A*-C including English and maths measure).
  • 68.8% of high attainers were entered for the EBacc and 55% achieved it. The entry rate is up 3.8 percentage points and the success rate up 2.9 points compared with 2013. Sixty-seven schools entered 100% of their high attainers, but only five schools managed 100% success. Thirty-seven schools entered no high attainers at all and 53 had no successful high attainers.
  • 85.6% of high attainers made at least the expected progress in English and 84.7% did so in maths. Both are down on 2013 but much more so in maths (3.1 percentage points) than in English (0.6 points).
  • In 108 schools every high attainer made the requisite progress in English. In 99 schools the same was true of maths in 99 schools. Only 21 schools managed 100% success in both English and maths. At the other extreme there were seven schools in which 50% or fewer made expected progress in both English and maths. Several schools recording 50% or below in either English or maths did significantly better with their middle attainers.
  • In sponsored academies one in four high attainers do not make the expected progress in maths and one in five do not do so in English. In free schools one in every five high attainers falls short in English as do one in six in maths.

At KS5:

  • 11.9% of students at state-funded schools and colleges achieved AAB grades at A level or higher, with at least two in facilitating subjects. This is a slight fall compared with the 12.1% that did so in 2013. The best-performing state institution had a success rate of 83%.
  • 14.1% of A levels taken in selective schools in 2014 were graded A* and 41.1% were graded A* or A. In selective schools 26.1% of the cohort achieved AAA or higher and 32.3% achieved AAB or higher with at least two in facilitating subjects.
  • Across all schools, independent as well as state-funded, the proportion of students achieving three or more A level grades at A*/A is falling and the gap between the success rates of boys and girls is increasing.
  • Boys are more successful than girls on three of the four high attainment measures, the only exception being the least demanding (AAB or higher in any subjects).
  • The highest recorded A level point score per A level student in a state-funded institution in 2014 is 1430.1, compared with an average of 772.7. The lowest is 288.4. The highest APS per A level entry is 271.1 compared with an average of 211.2. The lowest recorded is 108.6.

Disadvantaged high attainers:

  • On the majority of the KS4 headline measures gaps between FSM and non-FSM performance are increasing, even when the 2013 methodology is applied to control for the impact of the reforms affecting comparability. Very limited improvement has been made against any of the five headline measures between 2011 and 2014. It seems that the pupil premium has had little impact to date on either attainment or progress. Although no separate information is forthcoming about the performance of disadvantaged high attainers, it is highly likely that excellence gaps are equally unaffected.

.

Definitions and comparability issues 

Definitions

The Secondary and 16-18 Tables take very different approaches, since the former deals exclusively with high attainers while the latter concentrates exclusively on high attainment.

The Secondary Tables define high attainers according to their prior attainment on end of KS2 tests. Most learners in the 2014 GCSE cohort will have taken these five years previously, in 2009.

The new supporting documentation describes the distinction between high, middle and low attainers thus:

  • low attaining = those below level 4 in the key stage 2 tests
  • middle attaining = those at level 4 in the key stage 2 tests
  • high attaining = those above level 4 in the key stage 2 tests.

Last year the equivalent statement added:

‘To establish a pupil’s KS2 attainment level, we calculated the pupil’s average point score in national curriculum tests for English, maths and science and classified those with a point score of less than 24 as low; those between 24 and 29.99 as middle, and those with 30 or more as high attaining.’

This is now missing, but the methodology is presumably unchanged.

It means that high attainers will tend to be ‘all-rounders’, whose performance is at least middling in each assessment. Those who are exceptionally high achievers in one area but poor in others are unlikely to qualify.

There is nothing in the Secondary Tables or the supporting SFRs about high attainment, such as measures of GCSE achievement at grades A*/A.

By contrast, the 16-18 Tables do not distinguish high attainers, but do deploy a high attainment measure:

‘The percentage of A level students achieving grades AAB or higher in at least two facilitating subjects’

Facilitating subjects include:

‘biology, chemistry, physics, mathematics, further mathematics, geography, history, English literature, modern and classical languages.’

The supporting documentation says:

‘Students who already have a good idea of what they want to study at university should check the usual entry requirements for their chosen course and ensure that their choices at advanced level include any required subjects. Students who are less sure will want to keep their options open while they decide what to do. These students might want to consider choosing at least two facilitating subjects because they are most commonly required for entry to degree courses at Russell Group universities. The study of A levels in particular subjects does not, of course, guarantee anyone a place. Entry to university is competitive and achieving good grades is also important.’

The 2013 Tables also included percentages of students achieving three A levels at grades AAB or higher in facilitating subjects, but this has now been dropped.

The Statement of Intent for the 2014 Tables explains:

‘As announced in the government’s response to the consultation on 16-19 accountability earlier this year, we intend to maintain the AAB measure in performance tables as a standard of academic rigour. However, to address the concerns raised in the 16-19 accountability consultation, we will only require two of the subjects to be in facilitating subjects. Therefore, the indicator based on three facilitating subjects will no longer be reported in the performance tables.’

Both these measures appear in SFR03/15, alongside two others:

  • Percentage of students achieving 3 A*-A grades or better At A level or applied single/double award A level.
  • Percentage of students achieving grades AAB or better at A level or applied single/double award A level.

Comparability Issues 

When it comes to analysis of the Secondary Tables, comparisons with previous years are compromised by changes to the way in which performance is measured.

Both SFRs carry an initial warning:

‘Two major reforms have been implemented which affect the calculation of key stage 4 (KS4) performance measures data in 2014:

  1. Professor Alison Wolf’s Review of Vocational Education recommendations which:
  • restrict the qualifications counted
  • prevent any qualification from counting as larger than one GCSE
  • cap the number of non-GCSEs included in performance measures at two per pupil
  1. An early entry policy to only count a pupil’s first attempt at a qualification.’

SFR02/15 explains that some data has been presented ‘on two alternative bases’:

  • Using the 2014 methodology with the changes above applied and
  • Using a proxy 2013 methodology where the effect of these two changes has been removed.

It points out that more minor changes have not been accounted for, including the removal of unregulated IGCSEs, the application of discounting across different qualification types, the shift to linear GCSE formats and the removal of the speaking and listening component from English.

Moreover, the proxy measure does not:

‘…isolate the impact of changes in school behaviour due to policy changes. For example, we can count best entry results rather than first entry results but some schools will have adjusted their behaviours according to the policy changes and stopped entering pupils in the same patterns as they would have done before the policy was introduced.’

Nevertheless, the proxy is the best available guide to what outcomes would have been had the two reforms above not been introduced. Unfortunately, it has been applied rather sparingly.

Rather than ignore trends completely, this post includes information about changes in high attainers’ GCSE performance compared with previous years, not least so readers can see the impact of the changes that have been introduced.

It is important that we do not allow the impact of these changes to be used as a smokescreen masking negligible improvement or even declines in national performance on key measures.

But we cannot escape the fact that the 2014 figures are not fully comparable with those for previous years. Several of the tables in SFR06/2015 carry a warning in red to this effect (but not those in SFR 02/2015).

A few less substantive changes also impact slightly on the comparability of A level results: the withdrawal of January examinations and ‘automatic add back’ of students whose results were deferred from the previous year because they had not completed their 16-18 study programme.

.

Secondary outcomes

. 

The High Attainer Population 

The Secondary Performance Tables show that there were 172,115 high attainers from state-funded schools within the relevant cohort in 2014, who together account for 32.3% of the entire state-funded school cohort.

This is some 2% fewer than the 175,797 recorded in 2013, which constituted 32.4% of that year’s cohort.

SFR02/2015 provides information about the incidence of high, middle and low attainers by school type and gender.

Chart 1, below, compares the proportion of high attainers by type of school, showing changes since 2011.

The high attainer population across all state-funded mainstream schools has remained relatively stable over the period and currently stands at 32.9%. The corresponding percentage in LA-maintained mainstream schools is slightly lower: the difference is exactly two percentage points in 2014.

High attainers constitute only around one-fifth of the student population of sponsored academies, but close to double that in converter academies. The former percentage is relatively stable but the latter has fallen by some nine percentage points since 2011, presumably as the size of this sector has increased.

The percentage of high attainers in free schools is similar to that in converter academies but has fluctuated over the three years for which data is available. The comparison between 2014 and previous years will have been affected by the inclusion of UTCs and studio schools prior to 2014.

.

HA sec1

*Pre-2014 includes UTCs and studio schools; 2014 includes free schools only

Chart 1: Percentage of high attainers by school type, 2011-2014

. 

Table 1 shows that, in each year since 2011, there has been a slightly higher percentage of female high attainers than male, the gap varying between 0.4 percentage points (2012) and 1.8 percentage points (2011).

The percentage of high-attaining boys in 2014 is the lowest it has been over this period, while the percentage of high attaining girls is slightly higher than it was in 2013 but has not returned to 2011 levels.

Year Boys Girls
2014 32.1 33.7
2013 32.3 33.3
2012 33.4 33.8
2011 32.6 34.4

Table 1: Percentage of high attainers by gender, all state-funded mainstream schools 2011-14

Table 2 shows that the percentage of high attainers in selective schools is almost unchanged from 2013, at just under 89%. This compares with almost 31% in comprehensive schools, unchanged from 2013, and 21% in modern schools, the highest it has been over this period.

The 11.2% of learners in selective schools who are middle attainers remind us that selection by ability through 11-plus tests gives a somewhat different sample than selection exclusively on the basis of KS2 attainment.

. 

Year Selective Comprehensive Modern
2014 88.8 30.9 21.0
2013 88.9 30.9 20.5
2012 89.8 31.7 20.9
2011 90.3 31.6 20.4

Table 2: Percentage of high attainers by admissions practice, 2011-14

The SFR shows that these middle attainers in selective schools are less successful than their high attaining peers, and slightly less successful than high attainers in comprehensives, but they are considerably more successful than middle attaining learners in comprehensive schools.

For example, in 2014 the 5+ A*-C grades including English and maths measure is achieved by:

  • 97.8% of high attainers in selective schools
  • 92.2% of high attainers in comprehensive schools
  • 88.1% of middle attainers in selective schools and
  • 50.8% of middle attainers in comprehensive schools.

A previous post ‘The Politics of Selection: Grammar schools and disadvantage’ (November 2014) explored how some grammar schools are significantly more selective than others – as measured by the percentage of high attainers within their GCSE cohorts – and the fact that some comprehensives are more selective than some grammar schools.

This is again borne out by the 2014 Performance Tables, which show that 10 selective schools have a cohort consisting entirely of high attainers, the same as in 2013. Eighty-nine selective schools have a high attainer population of 90% or more.

However, five are at 70% or below, with the lowest – Dover Grammar School for Boys – registering only 52% high attainers.

By comparison, comprehensives such as King’s Priory School, North Shields and Dame Alice Owen’s School, Potters Bar record 86% and 77% high attainers respectively. 

There is also huge variation in modern schools, from Coombe Girls’ in Kingston, at 45%, just seven percentage points shy of the lowest recorded in a selective school, to The Ellington and Hereson School, Ramsgate, at just 4%.

Two studio colleges say they have no high attainers at all, while 96 schools have 10% or fewer. A significant proportion of these are academies located in rural and coastal areas.

Even though results are suppressed where there are too few high attainers, it is evident that these small cohorts perform very differently in different schools.

Amongst those with a high attainer population of 10% or fewer, the proportion achieving:

  • 5+ A*-C grades including English and maths varies from 44% to100%
  • EBacc ranges from 0% to 89%
  • expected progress in English varies between 22% and 100% and expected progress in maths between 27% and 100%. 

5+ GCSEs (or equivalent) at A*-C including GCSEs in English and maths 

The Tables show that:

  • 92.8% of high attainers in state-funded schools achieved five or more GCSEs (or equivalent) including GCSEs in English and maths. This compares with 56.6% of all learners. Allowing of course for the impact of 2014 reforms, the latter is a full four percentage points down on the 2013 outcome. By comparison, the outcome for high attainers is down 1.9 percentage points, slightly less than half the overall decline. Roughly one in every fourteen high attainers fails to achieve this benchmark.
  • 340 schools achieve 100% on this measure, significantly fewer than the 530 that did so in 2013 and the 480 managing this in 2012. In 2013, 14 schools registered 67% or fewer high attainers achieving this outcome, whereas in 2014 this number has increased substantially, to 57 schools. Five schools record 0%, including selective Bourne Grammar School, Lincolnshire, hopefully because of their choice of IGCSEs. Six more are at 25% or lower.

. 

A*-C grades in GCSE English and maths 

The Tables reveal that:

  • 93.8% of high attainers in state-funded schools achieved A*-C grades in GCSE English and maths, compared with 58.9% of all pupils. The latter percentage is down by 2.4 percentage points but the former has fallen by only 1.3 percentage points. Roughly one in 16 high attainers fails to achieve this measure.
  • In 2014 the number of schools with 100% of high attainers achieving this measure has fallen to some 470, 140 fewer than in 2013 and 60 fewer than in 2012. There were 38 schools recording 67% or lower, a significant increase compared with 12 in 2013 and 18 in 2012. Of these, four are listed at 0% (Bourne Grammar is at 1%) and five more are at 25% or lower.
  • Amongst the 38 schools recording 67% or lower, five return a higher success rate for their middle attainers than for their high attainers. Four of these are the same that do so on the 5+ A*-C measure above. They are joined by Tong High School. 

Entry to and achievement of the EBacc 

The Tables indicate that:

  • 68.8% of high attainers in state-funded schools were entered for all EBacc subjects and 55.0% achieved the EBacc. The entry rate is up by 3.8 percentage points compared with 2013, and the success rate is up by 2.9 percentage points. By comparison, 31.5% of middle attainers were entered (up 3.7 points) and 12.7% passed (up 0.9 points). Between 2012 and 2013 the entry rate for high attainers increased by 19 percentage points, so the rate of improvement has slowed significantly. Given the impending introduction of the Attainment 8 measure, commitment to the EBacc is presumably waning.
  • Thirty-seven schools entered no high attainers for the EBacc, compared with 55 in 2013 and 186 in 2012. Only 53 schools had no high attainers achieving the EBacc, compared with 79 in 2013 and 235 in 2012. Of these 53, 11 recorded a positive success rate for their middle attainers, though the difference was relatively small in all cases.

At least 3 Levels of Progress in English and maths

The Tables show that:

  • Across all state-funded schools 85.6% of high attainers made at least the expected progress in English while 84.7% did so in maths. The corresponding figures for middle attainers are 70.2% in English and 65.3% in maths. Compared with 2013, the percentages for high attainers are down 0.6 percentage points in English and down 3.1 percentage points in maths, presumably because the first entry only rule has had more impact in the latter. Even allowing for the depressing effect of the changes outlined above, it is unacceptable that more than one in every seven high attainers fails to make the requisite progress in each of these core subjects, especially when the progress expected is relatively undemanding for such students.
  • There were 108 schools in which every high attainer made at least the expected progress in English, exactly the same as in 2013. There were 99 schools which achieved the same outcome in maths, down significantly from 120 in 2013. In 2013 there were 36 schools which managed this in both English in maths, but only 21 did so in 2014.
  • At the other extreme, four schools recorded no high attainers making the expected progress in English, presumably because of their choice of IGCSE. Sixty-five schools were at or below 50% on this measure. In maths 67 schools were at or below 50%, but the lowest recorded outcome was 16%, at Oasis Academy, Hextable.
  • Half of the schools achieving 50% or less with their high attainers in English or maths also returned better results with middle attainers. Particularly glaring differentials in English include Red House Academy (50% middle attainers and 22% high attainers) and Wingfield Academy (73% middle attainers; 36% high attainers). In maths the worst examples are Oasis Academy Hextable (55% middle attainers and 16% high attainers), Sir John Hunt Community Sports College (45% middle attainers and 17% high attainers) and Roseberry College and Sixth Form (now closed) (49% middle attainers and 21% high attainers).

Comparing achievement of these measures by school type and admissions basis 

SFR02/2015 compares the performance of high attainers in different types of school on each of the five measures discussed above. This data is presented in Chart 2 below.

.

HA sec2 

Chart 2: Comparison of high attainers’ GCSE performance by type of school, 2014

.

It shows that:

  • There is significant variation on all five measures, though these are more pronounced for achievement of the EBacc, where there is a 20 percentage point difference between the success rates in sponsored academies (39.2%) and in converter academies (59.9%).
  • Converter academies are the strongest performers across the board, while sponsored academies are consistently the weakest. LA-maintained mainstream schools out-perform free schools on four of the five measures, the only exception being expected progress in maths.
  • Free schools and converter academies achieve stronger performance on progress in maths than on progress in English, but the reverse is true in sponsored academies and LA-maintained schools.
  • Sponsored academies and free schools are both registering relatively poor performance on the EBacc measure and the two progress measures.
  • One in four high attainers in sponsored academies fails to make the requisite progress in maths while one in five fail to do so in English. Moreover, one in five high attainers in free schools fails to make the expected progress in English and one in six in maths. This is unacceptably low.

Comparisons with 2013 outcomes show a general decline, with the exception of EBacc achievement.

This is particularly pronounced in sponsored academies, where there have been falls of 5.2 percentage points on 5+ A*-Cs including English and maths, 5.7 points on A*-C in English and maths and 4.7 points on expected progress in maths. However, expected progress in English has held up well by comparison, with a fall of just 0.6 percentage points.

Progress in maths has declined more than progress in English across the board. In converter academies progress in maths is down 3.1 points, while progress in English is down 1.1 points. In LA-maintained schools, the corresponding falls are 3.4 and 0.4 points respectively.

EBacc achievement is up by 4.5 percentage points in sponsored academies, 3.1 points in LA-maintained schools and 1.8 points in converter academies.

.

Comparing achievement of these measures by school admissions basis 

SFR02/2015 compares the performance of high attainers in selective, comprehensive and modern schools on these five measures. Chart 3 illustrates these comparisons.

.

HA sec3

Chart 3: Comparison of high attainers’ GCSE performance by school admissions basis, 2014

.

It is evident that:

  • High attainers in selective schools outperform those in comprehensive schools on all five measures. The biggest difference is in relation to EBacc achievement (21.6 percentage points). There is a 12.8 point advantage in relation to expected progress in maths and an 8.7 point advantage on expected progress in English.
  • Similarly, high attainers in comprehensive schools outperform those in modern schools. They enjoy a 14.7 percentage point advantage in relation to achievement of the EBacc, but, otherwise, the differences are between 1.6 and 3.5 percentage points.
  • Hence there is a smaller gap, by and large, between the performance of high attainers in modern and comprehensive schools respectively than there is between high attainers in comprehensive and selective schools respectively.
  • Only selective schools are more successful in achieving expected progress in maths than they are in English. It is a cause for some concern that, even in selective schools, 6.5% of pupils are failing to make at least three levels of progress in English.

Compared with 2013, results have typically improved in selective schools but worsened in comprehensive and modern schools. For example:

  • Achievement of the 5+ GCSE measure is up 0.5 percentage points in selective schools but down 2.3 points in comprehensives and modern schools.
  • In selective schools, the success rate for expected progress in English is up 0.5 points and in maths it is up 0.4 points. However, in comprehensive schools progress in English and maths are both down, by 0.7 points and 3.5 points respectively. In modern schools, progress in English is up 0.3 percentage points while progress in maths is down 4.1 percentage points.

When it comes to EBacc achievement, the success rate is unchanged in selective schools, up 3.1 points in comprehensives and up 5 points in modern schools.

. 

Other measures

The Secondary Performance Tables also provide information about the performance of high attainers on several other measures, including:

  • Average Points Score (APS): Annex B of the Statement of Intent says that, as in 2013, the Tables will include APS (best 8) for ‘all qualifications’ and ‘GCSEs only’. At the time of writing, only the former appears in the 2014 Tables. For high attainers, the APS (best 8) all qualifications across all state-funded schools is 386.2, which compares unfavourably with 396.1 in 2013. Four selective schools managed to exceed 450 points: Pate’s Grammar School (455.1); The Tiffin Girls’ School (452.1); Reading School (451.4); and Colyton Grammar School (450.6). The best result in 2013 was 459.5, again at Colyton Grammar School. At the other end of the table, only one school returns a score of under 250 for their high attainers, Pent Valley Technology College (248.1). The lowest recorded score in 2013 was significantly higher at 277.3.
  • Value Added (best 8) prior attainment: The VA score for all state-funded schools in 2014 is 1000.3, compared with 1001.5 in 2013. Five schools returned a result over 1050, whereas four did so in 2013. The 2014 leaders are: Tauheedul Islam Girls School (1070.7); Yesodey Hatorah Senior Girls School (1057.8); The City Academy Hackney (1051.4); The Skinner’s School (1051.2); and Hasmonean High School (1050.9). At the other extreme, 12 schools were at 900 or below, compared with just three in 2013. The lowest performer on this measure is Hull Studio School (851.2). 
  • Average grade: As in the case of APS, the average grade per pupil per GCSE has not yet materialised. The average grade per pupil per qualification is supplied. Five selective schools return A*-, including Henrietta Barnett, Pate’s, Reading School, Tiffin Girls and Tonbridge Grammar. Only Henrietta Barnett and Pate’s managed this in 2013.
  • Number of exam entries: Yet again we only have number of entries for all qualifications and not for GCSE only. The average number of entries per high attainer across state-funded schools is 10.4, compared with 12.1 in 2013. This 1.7 reduction is smaller than for middle attainers (down 2.5 from 11.4 to 8.9) and low attainers (down 3.7 from 10.1 to 6.4). The highest number of entries per high attainer was 14.2 at Gable Hall School and the lowest was 5.9 at The Midland Studio College Hinkley.

16-18: A level outcomes

.

A level grades AAB or higher in at least two facilitating subjects 

The 16-18 Tables show that 11.9% of students in state-funded schools and colleges achieved AAB+ with at least two in facilitating subjects. This is slightly lower than the 12.1% recorded in 2013.

The best-performing state-funded institution is a further education college, Cambridge Regional College, which records 83%. The only other state-funded institution above 80% is The Henrietta Barnett School. At the other end of the spectrum, some 443 institutions are at 0%.

Table 3, derived from SFR03/2015, reveals how performance on this measure has changed since 2013 for different types of institution and, for schools with different admission arrangements.

.

2013 2014
LA-maintained school 11.4 11.5
Sponsored academy 5.4 5.3
Converter academy 16.4 15.7
Free school* 11.3 16.4
Sixth form college 10.4 10
Other FE college 5.8 5.7
 
Selective school 32.4 32.3
Comprehensive school 10.7 10.5
Modern school 2 3.2

.

The substantive change for free schools will be affected by the inclusion of UTCs and studio schools in that line in 2013 and the addition of city technology colleges and 16-19 free schools in 2014.

Otherwise the general trend is slightly downwards but LA-maintained schools have improved very slightly and modern schools have improved significantly.

.

Other measures of high A level attainment

SFR03/15 provides outcomes for three other measures of high A level attainment:

  • 3 A*/A grades or better at A level, or applied single/double award A level
  • Grades AAB or better at A level, or applied single/double award A level
  • Grades AAB or better at A level all of which are in facilitating subjects.

Chart 4, below, compares performance across all state-funded schools and colleges on all four measures, showing results separately for boys and girls.

Boys are in the ascendancy on three of the four measures, the one exception being AAB grades or higher in any subjects. The gaps are more substantial where facilitating subjects are involved.

.

HA sec4

Chart 4: A level high attainment measures by gender, 2014

.

The SFR provides a time series for the achievement of the 3+ A*/A measure, for all schools – including independent schools – and colleges. The 2014 success rate is 12.0%, down 0.5 percentage points compared with 2013.

The trend over time is shown in Chart 5 below. This shows how results for boys and girls alike are slowly declining, having reached their peak in 2010/11. Boys established a clear lead from that year onwards.

As they decline, the lines for boys and girls are steadily diverging since girls’ results are falling more rapidly. The gap between boys and girls in 2014 is 1.3 percentage points.

.

HA sec5

Chart 5: Achievement of 3+ A*/A grades in independent and state-funded schools and in colleges, 2006-2014

.

Chart 6, compares performance on the four different measures by institutional type. It shows a similar pattern across the piece.

Success rates tend to be highest in either converter academies or free schools, while sponsored academies and other FE institutions tend to bring up the rear. LA-maintained schools and sixth form colleges lie midway between.

Converter academies outscore free schools when facilitating subjects do not enter the equation, but the reverse is true when they do. There is a similar relationship between sixth form colleges and LA-maintained schools, but it does not quite hold with the final pair.

. 

HA sec6 

Chart 6: Proportion of students achieving different A level high attainment measures by type of institution, 2014

.

Chart 7 compares performance by admissions policy in the schools sector on the four measures. Selective schools enjoy a big advantage on all four. More than one in four selective school students achieving at least 3 A grades and almost one in 3 achieves AAB+ with at least two in facilitating subjects.

There is a broadly similar relationship across all the measures, in that comprehensive schools record roughly three times the rates achieved in modern schools and selective schools manage roughly three times the success rates in comprehensive schools. 

. 

HA sec7 

Chart 7: Proportion of students achieving different A level high attainment measures by admissions basis in schools, 2014

 .

Other Performance Table measures 

Some of the other measures in the 16-18 Tables are relevant to high attainment:

  • Average Point Score per A level student: The APS per student across all state funded schools and colleges is 772.7, down slightly on the 782.3 recorded last year. The highest recorded APS in 2014 is 1430.1, by Colchester Royal Grammar School. This is almost 100 ahead of the next best school, Colyton Grammar, but well short of the highest score in 2013, which was 1650. The lowest APS for a state-funded school in 2014 is 288.4 at Hartsdown Academy, which also returned the lowest score in 2013. 
  • Average Point Score per A level entry: The APS per A level entry for all state-funded institutions is 211.2, almost identical to the 211.3 recorded in 2013. The highest score attributable to a state-funded institution is 271.1 at The Henrietta Barnett School. This is very slightly slower than the 271.4 achieved by Queen Elizabeth’s Barnet in 2013. The lowest is 108.6, again at Hartsdown Academy, which exceeds the 2013 low of 97.7 at Appleton Academy. 
  • Average grade per A level entry: The average grade across state-funded schools and colleges is C. The highest average grade returned in the state-funded sector is A at The Henrietta Barnett School, Pate’s Grammar School, Queen Elizabeth’s Barnet and Tiffin Girls School. In 2013 only the two Barnet schools achieved the same outcome. At the other extreme, an average U grade is returned by Hartsdown Academy, Irlam and Cadishead College and Swadelands School. 

SFR06/2015 also supplies the percentage of A* and A*/A grades by type of institution and schools’ admissions arrangements. The former is shown in Chart 8 and the latter in Chart 9 below.

The free school comparisons are affected by the changes to this category described above.

Elsewhere the pattern is rather inconsistent. Success rates at A* exceed those set in 2012 and 2013 in LA-maintained schools, sponsored academies, sixth form colleges and other FE institutions. Meanwhile, A*/A grades combined are lower than both 2012 and 2013 in converter academies and sixth form colleges.

.

HA sec8

Chart 8: A level A* and A*/A performance by institutional type, 2012 to 2014

. 

Chart 9 shows A* performance exceeding the success rates for 2012 and 2013 in all three sectors.

When both grades are included, success rates in selective schools have returned almost to 2012 levels following a dip in 2013, while there has been little change across the three years in comprehensive schools and a clear improvement in modern schools, which also experienced a dip last year.

HA sec9

Chart 9: A level A* and A*/A performance in schools by admissions basis, 2012 to 2014.

 .

Disadvantaged high attainers 

There is nothing in either of the Performance Tables or the supporting SFRs to enable us to detect changes in the performance of disadvantaged high attainers relative to their more advantaged peers.

I dedicated a previous post to the very few published statistics available to quantify the size of these excellence gaps and establish if they are closing, stable or widening.

There is continuing uncertainty whether this will be addressed under the new assessment and accountability arrangements to be introduced from 2016.

Although results for all high attainers appear to be holding up better than those for middle and lower attainers, the evidence suggests that FSM and disadvantaged gaps at lower attainment levels are proving stubbornly resistant to closure.

Data from SFR06/2015 is presented in Charts 10-12 below.

Chart 10 shows that, when the 2014 methodology is applied, three of the gaps on the five headline measures increased in 2014 compared with 2013.

That might have been expected given the impact of the changes discussed above but, if the 2013 methodology is applied, so stripping out much (but not all) of the impact of these reforms, four of the five headline gaps worsened and the original three are even wider.

This seems to support the hypothesis that the reforms themselves are not driving this negative trend, athough Teach First has suggested otherwise.

.

HA sec10

Chart 10: FSM gaps for headline GCSE measures, 2013-2014

.

Chart 11 shows how FSM gaps have changed on each of these five measures since 2011. Both sets of 2014 figures are included.

Compared with 2011, there has been improvement on two of the five measures, while two or three have deteriorated, depending which methodology is applied for 2014.

Since 2012, only one measure has improved (expected progress in English) and that by slightly more or less than 1%, according to which 2014 methodology is selected.

Deteriorations have been small however, suggesting that FSM gaps have been relatively stable over this period, despite their closure being a top priority for the Government, backed up by extensive pupil premium funding.

.

HA sec11

Chart 11: FSM/other gaps for headline GCSE measures, 2011 to 2014.

.

Chart 12 shows a slightly more positive pattern for the gaps between disadvantaged learners (essentially ‘ever 6 FSM’ and looked after children) and their peers.

There have been improvements on four of the five headline measures since 2011. But since 2012, only one or two of the measures has improved, according to which 2014 methodology is selected. Compared with 2013, either three or four of the 2014 headline measures are down.

The application of the 2013 methodology in 2014, rather than the 2014 methodology, causes all five of the gaps to increase, so reinforcing the point in bold above.

It is unlikely that this pattern will be any different at higher attainment levels, but evidence to prove or disprove this remains disturbingly elusive.

.

HA sec12

Chart 12: Disadvantaged/other gaps for headline GCSE measures, 2011 to 2014

.

Taken together, this evidence does not provide a ringing endorsement of the Government’s strategy for closing these gaps.

There are various reasons why this might be the case:

  • It is too soon to see a significant effect from the pupil premium or other Government reforms: This is the most likely defensive line, although it begs the question why more urgent action was/is discounted.
  • Pupil premium is insufficiently targeted at the students/school that need it most: This is presumably what underlies the Fair Education Alliance’s misguided recommendation that pupil premium funding should be diverted away from high attaining disadvantaged learners towards their lower attaining peers.
  • Schools enjoy too much flexibility over how they use the pupil premium and too many are using it unwisely: This might point towards more rigorous evaluation, tighter accountability mechanisms and stronger guidance.
  • Pupil premium funding is too low to make a real difference: This might be advanced by institutions concerned at the impact of cuts elsewhere in their budgets.
  • Money isn’t the answer: This might suggest that the pupil premium concept is fundamentally misguided and that the system as a whole needs to take a different or more holistic approach.

I have proposed a more targeted method of tackling secondary excellence gaps and simultaneously strengthening fair access, where funding topsliced from the pupil premium is fed into personal budgets for disadvantaged high attainers.

These would meet the cost of coherent, long-term personalised support programmes, co-ordinated by their schools and colleges, which would access suitable services from a ‘managed market’ of suppliers.

.

Conclusion

This analysis suggests that high attainers, particularly those in selective schools, have been relatively less affected by the reforms that have depressed GCSE results in 2014.

While we should be thankful for small mercies, three issues are of particular concern:

  • There is a stubborn and serious problem with the achievement of expected progress in both English and maths. It cannot be acceptable that approximately one in seven high attainers fails to make three levels of progress in each core subject when this is a relatively undemanding expectation for those with high prior attainment. This issue is particularly acute in sponsored academies where one in four or five high attainers are undershooting their progress targets.
  • Underachievement amongst high attainers is prevalent in far too many state-funded schools and colleges. At KS4 there are huge variations in the performance of high-attaining students depending on which schools they attend. A handful of schools achieve better outcomes with their middle attainers than with their high attainers. This ought to be a strong signal, to the schools as well as to Ofsted, that something serious is amiss.
  • Progress in closing KS4 FSM gaps continues to be elusive, despite this being a national priority, backed up by a pupil premium budget of £2.5bn a year. In the absence of data about the performance of disadvantaged high attainers, we can only assume that this is equally true of excellence gaps.

.

GP

February 2015

16-19 Maths Free Schools Revisited: Oddyssean Edition

This is the second edition of a post that marks the opening of two university-sponsored 16-19 maths free schools by taking a fresh look at the wider programme that spawned them.

Courtesy of Andrew J Hanson Indiana University

Courtesy of Andrew J Hanson Indiana University

I have revised the text to reflect substantive comments provided by Dominic Cummings through the oddyseanproject Twitter feed. Cummings was political adviser to former Secretary of State for Education Michael Gove until January 2014. He was instigator and champion of the maths free schools programme.

I fee obliged to point out that the inclusion of these comments does not constitute his endorsement or approval of the text. I have reserved the right to part company with him on matters of interpretation (rather than matters of fact) and have signalled where instances occur.

The post scrutinises developments since the publication of ‘A Progress Report on 16-19 Maths Free Schools’ (March 2013), building on the foundations within ‘The Introduction in England of Selective 16-19 Maths Free Schools’ (November 2011).

The broad structure of the post is as follows:

  • A description of the genesis of the programme and a summary of developments up to March 2013.
  • The subsequent history of the programme, from March 2013 to the present day. This reviews efforts to recruit more university sponsors into the programme – and to resist the publication of information showing which had submitted expressions of interest and, subsequently, formal proposals.
  • An assessment of the prospects for the programme at this point and for wider efforts to expand and remodel England’s national maths talent pipeline.

Since many readers will be interested in some of these sections but not others, I have included direct links to the main text from the first word of each bullet point above.

 

Genesis and early developments

Capital investment to support the programme was confirmed in the 2011 Autumn Statement, which referred to:

‘…an extra £600 million to fund 100 additional Free Schools by the end of this parliament. This will include new specialist maths Free Schools for 16-18 year olds, supported by strong university maths departments and academics’.

This followed an orchestrated sequence of stories fed to the media immediately prior to the Statement.

One source reported a plan to establish 12 such schools in major cities by the end of the Parliament (Spring 2015) ‘before the model is expanded nationwide’. These would:

‘…act as a model for similar institutions specialising in other subjects’.

Another confirmed the number of institutions, adding that there would be ‘…a special application process outside the regular free school application process…’

A third added that the project was viewed as an important part of the Government’s strategy for economic growth, suggesting that some of the schools:

‘…would offer pure maths, while others would combine the subject with physics, chemistry or computer sciences.’

Assuming provision for 12 schools at £6m a time, the Treasury had provided a capital budget of £72m available until 2015. It remains unclear whether this sum was ringfenced for university-sponsored maths schools or could be diverted into the wider free schools programme.

We now know that Cummings was behind the maths free schools project. But these original press briefings originated from the Treasury, showing that they were indeed committed to a 12-school programme within the lifetime of the Parliament.

 

 

The most recent edition of Cummings’ essay ‘Some thoughts on education and political priorities’ (2013) sets out the rationale for the programme:

‘We know that at the top end of the ability range, specialist schools, such as the famous Russian ‘Kolmogorov schools’…show that it is possible to educate the most able and interested pupils to an extremely high level…We should give this ~2% a specialist education as per Eton or Kolmogorov, including deep problem-solving skills in maths and physics.

The first English specialist maths schools, run by King’s College and Exeter University, have been approved by the Department for Education and will open in 2014. All of the pupils will be prepared for the maths ‘STEP’ paper that Cambridge requires for entry (or Oxford’s equivalent) – an exam that sets challenging problems involving unfamiliar ways of considering familiar  material, rather than the formulaic multi-step questions of A Level.’

Back in February 2012, TES reported that:

‘The DfE has hosted a consultation meeting on the new free schools with interested parties from the mathematical community in order to outline its plans.’

‘TES understands that officials within the Department for Education are now keen to establish the schools on the model of Kolmogorov, a boarding school that selects the brightest mathematicians in Russia.’

Andrey Kolmogorov courtesy of Svjo

Andrey Kolmogorov courtesy of Svjo

 

In fact, the meeting discussed a variety of international models and, on 20 February, Education Minister Nick Gibb answered a PQ thus:

‘Alex Cunningham: To ask the Secretary of State for Education when he expects the first free school specialising in mathematics for 16 to 18 year-olds to open; how many 16 to 18 year-olds he expects to enrol in free schools specialising in mathematics by 2015; with which universities he has discussed these free schools; and what guidance he plans to provide to people who wish to apply to open such a school.

Mr Gibb: We are developing proposals on how specialist maths schools for 16 to 18-year-olds might operate and will announce further details in due course. We are keen to engage with all those who have an interest to explore possible models and innovative ideas.’ (Col. 723W).

However, no proposals were published.

The minutes from King’s College London (KCL) Council meeting of 26 June 2012 reveal that:

‘Following approval by the Principal’s Central Team, the College was pursuing discussions with the Department for Education about sponsoring one of 12 specialist Maths schools for 16-18 year olds to be established with the support of university Mathematics departments. The initiative was intended to address national deficiencies in the subject and to promote a flow of highly talented students into university. In discussion, members noted that while the financial and reputational risks and the costs in management time needed to be carefully analysed, the project supported the College’s commitment to widening participation and had the potential to enhance the strengths of the Mathematics Department and the Department of Education and Professional Services, as well as addressing a national problem. The Council approved the College’s continued engagement with this initiative.’

By December 2012 KCL had announced that it would establish a maths free school, with both its maths and education departments involved. The school was scheduled to open in September 2014.

KCL confirmed that it had received from DfE a development grant plus a parallel outreach grant to support a programme for mathematically talented 14-16 year-olds, some of whom might subsequently attend the school.

The minutes of the University of Exeter Council meeting of 13 December 2012 record that:

‘As Council were aware, Exeter was going to be a partner in an exciting regional development to set up one of the first two Maths specialist schools with Exeter College. The other school would be led by King’s College London. This would cater for talented Maths students as a Free School with intake from four counties (Devon, Cornwall, Somerset and Dorset) with a planned total number of students of 120 by September 2017. The bid was submitted to the Department of Education on 11th December and the outcome would be announced in early January, with the school opening in 2014. It would be taught by Exeter College teachers with contributions from staff in pure and applied Maths in the College of Engineering, Mathematics and Physical Sciences (CEMPS), input from the Graduate School of Education and from CEMPS students as mentors and ambassadors. It was hoped that at least some of these talented students would choose to progress to the University. Council would be kept informed of the progress of the bid.’

In January 2013 a DfE press release announced approval of this second school. It would indeed have capacity for 120 students, with Monday-Thursday boarding provision for 20% (24 students), enabling it to recruit from across the four counties named above, so acting as a ‘regional centre of excellence’.

This project had also received a development grant – which we know was up to £300K – had agreement in principle to an outreach grant and also expected to open in September 2014.

There is also reference to plans for Met Office involvement with the School.

The press release repeats that:

‘The ultimate aim is to create a network of schools that operate across England which identify and nurture mathematical and scientific talent.’

A page added to DfE’s website in March 2013 invites further expressions of interest to open maths free schools in September 2014 and beyond.

Parallel Q and A, which has now been removed, made clear that development grants would not be available to new applicants:

‘Is there financial support available to develop our plans?

Not at the beginning. Once we have approved a proposal, we do offer some support to cover the costs of project management, and recruiting some staff before the school opens, in the same way we would for any Free School.’

This has subsequently been reversed (see below).

 

Progress since March 2013

 

The Hard Sell

While KCL and Exeter developed their plans, strenuous efforts were made to encourage other universities to participate in the programme.

A TES piece from May 2013, profiling the newly-appointed head of the KCL school, includes a quote from Alison Wolf – the prominent chair of the project group at KCL:

‘’The Brit School is a really good comparison,” she says. “When we were working on the new school and thinking about what to do, we’d look at their website.

“Maths is very glamorous if you’re a young mathematician, which is why they’ll do well when they are around other people who adore maths.”

The story adds that 16 schools are now planned rather than the original 12, but no source is attributed to this statement. Cummings says it is a mistake

 

 

It seems that the wider strategy at this stage was to convince other potential university sponsors that maths schools were an opportunity not to be missed, to imply that there was already substantial interest from prominent competitors, so encouraging them to climb on board for fear of missing the boat.

 

Playing the Fair Access Card

But there was soon an apparent change of tack. In June 2013, the Guardian reported that education minister Liz Truss had written to the heads of university maths departments to encourage bids.

‘As an incentive to open the new schools, universities will be allowed to fund them using budgets otherwise reserved for improving access to higher education for under-represented and disadvantaged groups….

Les Ebdon, director of Offa, said: “I’d be happy to see more university-led maths free schools because of the role they can play in helping able students from disadvantaged backgrounds access higher education.

“It is for individual universities and colleges to decide whether or not this is something they want to do, but Offa is supportive of anything that is targeted at under-represented groups and helps them to fulfil their potential.”

…According to Truss’s letter, Ebdon confirmed it would be “perfectly legitimate to allocate funding ringfenced for improving access for under-represented groups towards the establishment of such schools,” counting the spending as “widening access”.’

My initial post had pointed to the potential significance of this coupling of excellence and equity as early as November 2011:

‘It is not clear whether a fundamental purpose of these institutions is to support the Government’s drive towards greater social mobility through fair access to competitive universities. However, one might reasonably suggest it would be an oversight not to deploy them…encouraging institutions to give priority during the admissions process would be the likely solution.’

What appeared to be Ministers’ rather belated conversion to the merits of alignment with social mobility and fair access might have been interpreted as opportunism rather than a sincere effort to join together two parallel strands of Government policy, especially since it had not been identified as a central feature in either KCL’s or Exeter’s plans.

But Cummings reveals that such alignment was intended from the outset.

 

 

I can find nothing on Offa’s website confirming the statement that funding ringfenced for fair access might be allocated by universities to the development of maths free schools. There is no contemporary press notice and nothing in subsequent guidance on the content of access agreements. This begs the question whether Ebdon’s comments constitute official Offa advice.

I asked Cummings why it took so long to get the line from Ebdon and why that line wasn’t encapsulated in Offa guidance.

 

 

The Cummings view of the dysfunctionality of central government is well-known, but to have to wait nineteen months for a brief statement on a high-priority programme – with inevitably long lead times yet time-limited to the duration of the Parliament – must have been deeply frustrating.

It would seem that Offa had to be persuaded away from sympathy with the negative views Cummings attributes to so many vice chancellors – and that this required a personal meeting at ministerial level.

But this was a priority programme with strong ministerial backing.

 

 

One must draw one’s own private conclusions about the motivations and commitment of the key protagonists – I will not apportion blame.

The text of Truss’s letter is preserved online and the identical text appears within it:

‘I want to encourage other universities to consider whether they could run similar schools: selective, innovative and stretching our brightest and best young mathematicians. It is a logical extension of the role that dozens of universities have already played in sponsoring academies.

I also wanted to highlight to your colleagues that Professor Les Ebdon, Director of the Office for Fair Access, is enthusiastic about the role university led Maths Free Schools can have in encouraging more young people to go on to study maths at university, and to reap the benefits that brings. Professor Ebdon has also confirmed to me that he considers the sponsorship and development of Maths Free Schools as contributing to higher education ‘widening access’ activity, and that it would be perfectly legitimate to allocate funding ring-fenced for improving access for underrepresented groups towards the establishment of such schools.

Unlike our usual practice for Free Schools, there is no competitive application process for Maths Free Schools. Instead we ask interested universities to submit a short proposal setting out the key features of the school. These proposals need not be long: King’s and Exeter both submitted initial proposals that were around 12 pages…

[There follows a list of bullet points describing the content of these initial proposals, none of which address the admission of students from disadvantaged backgrounds.]

….Both King’s College and the University of Exeter had a number of detailed discussions with colleagues in the Department to develop and refine their proposals and we are always happy to work with universities to help them focus their plans before submitting a formal proposal. If we approve a proposal, we do then offer financial support to cover the costs of project management, and of recruiting some staff before the school opens, in the same we would for any free school.’

(By way of an aside, note that the final emboldened sentence in the quotation above corrects the statement in the Q and A mentioned above. It seems that maths free schools are now treated comparably with all other free school projects in this respect, even though the application process remains different.

The latest version of free school pre-opening guidance gives the sum available in Project Development Grant for 16-19 free schools as £0.25m.)

Going back to Offa, there are no conditions imposed by Ebdon in respect of admissions to the schools, which seems a little over-relaxed, given that they might well attract a predominantly advantaged intake. I wonder whether Ebdon was content to offer personal support but refused to provide official Offa endorsement.

 

 

In July 2013 the BBC reported a speech by Truss at the 2013 ACME Conference. Oddly, the speech is not preserved on the gov.uk site. According to the BBC:

“We want this movement to spread still further,” she told delegates.

“So we’re allowing universities to apply to sponsor new maths free schools through a fast-track, simplified procedure, without having to go through the normal competitive application process.

“These schools will not only improve standards in maths teaching, but will equip talented young people from low-income backgrounds with the skills they need to study maths at university.”

Mrs Truss said the Office for Fair Access had confirmed that, when universities contributed to the sponsorship or development of maths free schools, this would be considered as one of their activities to widen access to under-represented groups – and therefore as part of their access agreement.

“I hope that this is the start of a new network of world-class free schools, under the aegis of top universities, helping to prepare talented 16- to 19-year-olds from any and every background for the demands of university study.”

Note that Ebdon’s endorsement is now Offa’s.

Cummings’ essay remarks in a footnote:

‘Other maths departments were enthusiastic about the idea but Vice Chancellor offices were hostile because of the political fear of accusations of ‘elitism’. Hopefully the recent support of Les Ebdon for the idea will change this.’

A year on, we have no evidence that it has done so. Cummings comments

 

 

What that ‘not none’ amounts to – beyond references (reproduced later in this post) in KCL’s and Exeter’s access agreements – remains to be established for, as we shall see, it does not feature prominently in the priorities of either of their schools.

 

The Soft Sell

By the beginning of the following academic year, a more subtle strategy was adopted. The two schools-in-development launched a maths competition for teams from London and the South-West with prizes awarded by education ministers.

 

 

A November 2013 DfE press release marks the ceremony. Michael Gove is quoted:

‘We need specialist maths free schools like King’s College London (KCL) Maths School and Exeter Mathematics School. They will develop the talents of exceptional young mathematicians and ensure they can compete in the global race.’

The release continues:

‘The KCL and Exeter schools are the first to take advantage of a development grant made available by the Department for Education for the creation of university-led specialist maths free schools.’

The notes include a link to the 1 March webpage mentioned above for ‘Universities interested in developing their own maths free school’.

 

Publicity avoided

We now know that a Freedom of Information request had been submitted to DfE in October 2013, asking how many expressions of interest and firm proposals had been received, which institutions had submitted these and which proposals had been approved and rejected.

The source is an ICO Decision Notice published on 12 June 2014.

The request was initially rejected and this decision was upheld in January 2014 following an internal review. A complaint was immediately lodged with the Information Commissioner’s Office.

The Decision Notice records the Commissioner’s decision that public interest outweighs the case for withholding the information. Accordingly he directs that it should be released to the complainant within 35 calendar days of the date of the Notice (ie by 17 July 2014).

The Notice contains some interesting snippets:

  • ‘It has been the DfE’s experience that interested Heads of Maths have contacted it for further information before seeking to discuss the idea with their Vice Chancellor.’ There is no process for accepting formal expressions of interest.
  • There are…no fixed criteria against which all proposals are assessed.’
  • ‘The DfE confirmed that the application is and has always been the first formal stage of the maths free schools process and it has already stated publicly that it has received three applications from King’s College London, Exeter University and the University of Central Lancashire.’
  • ‘It [ie DfE] confirmed that funding arrangements were only confirmed for the development of maths free schools in February 2014 and many policy decisions on this issue have been shaped by the specifics of the two schools that are due to open soon. It expects the policy to develop even further as more maths free schools are approved.’
  • ‘The DfE explained that universities are extremely risk adverse when it comes to protecting their reputation and so do not want to be publically named until they have submitted an application. As such, if they are named at an earlier point it may make them pull out altogether and may make universities unwilling to approach the DfE with ideas.’
  • ‘Similarly, the DfE argued that if it were to release the reasons why one of the applications was rejected it would be likely to deter future interest as the university would not want the public criticism of its ideas. Given that the policy is driven by university interest, if all potential groups are deterred the policy will fail and students will not be able to enjoy the potential benefits.’

The Commissioner gave these arguments short shrift, pointing out the benefits of transparency for policy development and the encouragement of more successful applications.

The text does not say so explicitly, but one can imagine the Commissioner thinking  ‘given the low level of interest stimulated to date, you might at least try a more open strategy – what have you got to lose?’

It does seem unlikely that university heads of maths departments would submit speculative expressions of interest without internal clearance. Their approaches were presumably of the informal ‘sounding out’ variety. They would understand the shaky internal politics of failing to consult the corporate centre – not to mention their education faculties

The lack of specific and transparent assessment criteria does appear to have backfired. What guarantees might universities otherwise receive that their proposals would be judged objectively?

One can imagine the questions:

  • Is the scheme open to all universities, Russell Group or otherwise?
  • If not, what criteria must the host university satisfy?
  • What counts as a ‘strong mathematics department?’
  • Can projects be led by university departments of education, or only undertaken jointly (as at KCL)?

Without explicit and consistent answers one can readily understand why many universities would be disinclined to pursue the idea.

Cummings disagrees strongly with this suggestion

 

 

But I am still unconvinced. Personal experience of working with sceptical vice chancellors and their offices leads me to believe that some distinct parameters would have been better than none, provided that they were flexible parameters, in all the areas where ministers were genuinely flexible.

Some flagging up of ministerial preferences might also have been helpful, provided it was also made clear that ministers could be persuaded away from them by a strong enough bid with a different complexion.

Since ministers set so much store by the fair access dimension, and were acutely aware of the need to face down universities’ concerns about elitism, some explicit statement of the importance they attached to this dimension would not have gone amiss.

And the reference to bespoke solutions rings rather hollow when – as we shall see – the proposals from KCL and Exeter were so strikingly similar.

I suspect this difference of opinion boils down to ideology – our very different ideas about bureaucracy and how best to harness innovation. The point is moot in any case.

 

The reference to belated confirmation of funding arrangements – as recently as February 2014 – is intriguing. It cannot apply to capital funding, unless that was vired in extremis. I wondered whether it might relate to the parallel recurrent funding pot or simply the availability of project development grants.

The latter seems unlikely given the statement in the letter to HoDOMS, dated some eight months previously.

One suspects that there might have been internal difficulties in ringfencing sufficient recurrent funding to honour proposals as and when they were received. Some prospective bidders might have baulked on being told that their budget could not be confirmed until a later date.

But the eventual resolution of this issue a little over a year before the end of the spending round would be unlikely to have a significant impact on the number of successful bids, especially if unspent capital funding has to be surrendered by Spring 2015.

Cummings throws some light on this issue

 

 

It sounds as though there were internal pressures to integrate maths free schools into the 16-19 free schools programme, where levels of bureaucracy might have caused further delay

But these comments tend to play down the budgetary issue flagged up to the ICO. Although it might have been strictly correct that: ‘funding arrangements were only confirmed for the development of maths free schools in February 2014‘, the associated suggestion that this had been a significant factor holding up the approval of further projects seems rather more suspect.

 

Recent developments

In July 2014 the TES revealed that it had been the source of this FoI request.

 

 

But the story itself reveals little new, other than that:

‘Five further expressions of interest have been made but not yet yielded an application’

The sources of these EoIs are not listed, even though they must have been divulged to the paper by this point.

David Reynolds opines that:

‘Having a small number of schools doesn’t matter if we can get the knowledge from them around the system. So we need them to be excellent schools and we need to somehow get that knowledge around.’

A DfE statement concludes:

‘We continue to welcome applications and expressions of interest from universities and the first maths free schools, set up by two leading universities, will be opening in September.’

So we know there have been eight expressions of interest, three of them converted into firm proposals.

The receipt of the third proposal, from the University of Central Lancashire (UClan), is said to have been made public, but I can find no record of it in the lists of Wave 1 to 7 free school applications so far released, or anywhere else for that matter. (KCL and Exeter were both included in Wave 3.)

There is a reference in UCLAN’s 2013-14 access agreement dated 31 May 2012:

‘The University is currently consulting on the formation of a Maths Free School which would be run alongside its new Engineering Innovation Centre at the Preston Campus.’

Nothing is said about the plans in the access agreements for 2014-15 and 2015-16.

There is one further reference on the New Schools Network site to a:

‘Consultant engaged to carry out a feasibility study re a Maths Free School on behalf of the University of Central Lancashire (UCLan)’.

One assumes that this must be out-of-date, unless UCLan is considering a second bid.

Otherwise, a simple process of elimination tells us that UCLan’s proposal must have been rejected. The reason for this is now presumably known to TES, as are the sources of the five expressions of interest that were not converted into proposals. Why have they not published this information?

Perhaps they are waiting for DfE to place these details on its website but, at the time of writing – almost three months after the Decision Notice issued – it has not been uploaded.

Meanwhile, there are no further maths free school proposals in the most recent Wave 7 information relating to applications received by 9 May 2014.

The deadline for Wave 8 is imminent. That may well be the last on this side of the Election.

Cummings reveals that there is a fourth proposal in the pipeline which is not yet ready to be made public.

 

 

One assumes a September 2015 start and we must wait to see whether it catches Wave 8.

We discussed the relationship of this proposal to the evidence submitted to the ICO. We do not know whether it features among the five expressions of interest but it might be supernumary. Cummings is at pains to justify a cautious approach to FoI requests.

 

 

He is content to release details only at the point where development funding is committed.

So, assuming DfE is pursuing the same strategy, one can reasonably conclude that development funding has not yet been agreed for this fourth proposal. Although it has progressed beyond the status of an expression of interest, it is not yet an approved application.

Almost nine months has passed since Cummings left the Department, yet negotiations have not reached the point where development funding is confirmed. This must be a complex and sensitive negotation indeed! Perhaps there is a Big Fish on the end of this particular hook…or perhaps the host university has cold feet. We must wait and see.

A further feature published by the TES in October 2014 throws no fresh light on these matters, though it carries a quote by new Secretary of State Nicky Morgan, interviewed at the KCL School launch:

‘I think that some [universities] are clearly waiting to see how the King’s and Exeter schools go. Clearly there is a huge amount of effort required, but I think King’s will be enormously successful, and I am hoping they will be leading by example.’

That sounds suspiciously like tacit admission that there will be no new proposals before a General Election.

Another opinion, diametrically opposed to David Reynolds’ view, is contributed by the head of the school of education at Nottingham University who is also Deputy Chair of ACME:

‘I’m very supportive of more people doing more maths, but even if you have 12 schools, you are really scratching the surface,” said Andrew Noyes, head of the school of education at Nottingham University and a former maths teacher.

“These kinds of policy experiments are very nice and they’re beneficial for a certain number of young people, but they’re relatively cheap compared with providing high-quality maths education at every stage in every school.”’

So what are the prospects for the success of the KCL and Exeter Schools? The next section reviews the evidence so far in the public domain.

 

The KCL and Exeter Free Schools

 

KCL School

The KCL School opened in September 2014 with 68 students, against a planned admissions number of 60. The most recent TES article says that there were 130 applicants and nearly all of those successful were drawn from state schools.

However, another reliable source – a member of the governing body – says that only 85% (ie 58) are from maintained schools, so the independent sector is actually over-represented.

He adds that:

‘Many are from families where neither parent has attended university’

but that is not necessarily an indicator of disadvantage.

We also know that some 43% (29 students) were female, which is a laudable outcome.

The School is located in Lambeth Walk, some distance from KCL’s main campuses. The capital cost of refurbishing the School was seemingly £5m. It occupies two buildings and the main building is shared with a doctor’s surgery.

My March 2013 post summarised KCL’s plans, as revealed by material on the University’s site at that time, supplemented by the content of an information pack for potential heads which is no longer available online.

I have reproduced the main points below, to provide a baseline against which to judge the finished article.

  • The full roll will be 120, with an annual admission number of 60. Potential applicants must have at least 5 GCSE grades A*-C including A*/A in both maths and physics or maths and dual award science.
  • Other admissions criteria will probably include a school reference, ‘our judgement about how much difference attending the school will make to your future based on a number of factors, including the results from an interview’ and the results of a test of aptitude for problem-solving and mathematical thinking.
  • The headteacher information pack adds that ‘the school will also be committed to recruiting a significant proportion of students from socially disadvantaged backgrounds, and to an outreach programme… to further this objective.’
  • All students will take Maths, Further Maths and Physics A levels. They will be expected to take STEP papers and may take a further AS level (an FAQ suggests this will be an Extended Project). Every student will have a maths mentor, either an undergraduate or ‘a junior member of the maths department’.
  • They will also ‘continue with a broad general curriculum, including other sciences, social science, humanities and languages, and have opportunities for sport and the visual and performing arts.’ Some of this provision will be ‘delivered through existing King’s facilities’. The provisional timetable assumes a 40-hour working week, including independent study.
  • The University maths department ‘will be closely involved in curriculum development’ and academics will have ‘regular timetabled contact’, potentially via masterclasses.
  • There will be strong emphasis on collaboration with partner schools. In the longer term, the school ‘intends to seek independent funding for a larger CPD programme associated with the school’s curriculum and pedagogy, and to offer it to a wide range of  schools and students, using school premises out of hours’.

At the time of writing, the KCL Maths School website does not have a working link to the admissions policy, although it can be found online.

As expected, 60 students will be admitted in September 2015. Minimum requirements are now

‘A or A* in GCSE Mathematics or in iGCSE Mathematics

Either an A or A* in GCSE Physics or iGCSE Physics, or an AA, A*A or A*A* in GCSE Science and GCSE Additional Science, or an A or A* in all three Physics modules contained within the GCSE Science, Additional Science and Further Additional Science qualifications; and

A*-C grade in 5 other GCSEs or other qualifications that count towards the Key Stage 4 performance tables compiled by the Department of Education, normally including English language.’

So the minimum requirement has been stiffened to at least seven GCSEs, or equivalent, including A*/A grades in maths and physics and at least a C in English language.

The application process does indeed include a reference, an aptitude test and an interview.

The test is based on KS3 national curriculum material up to Level 8, containing ‘routine and less familiar problems’. Some specimen questions are supplied.

The latest TES story says there are two interviews but this is wrong – there is one interview but two interview scores.

Cummings queries this point

 

 

I can no longer check the original admissions policy to establish whether there was exceptionally provision for two interviews for admission in 2014, but all the other material I have seen – including the admissions policy for 2015 – refers to a single interview.

One of the two scores is ‘to assess to what extent the school is likely to add value in terms of making a difference to [candidates’] future careers’ but there is no explicit reference to priority for disadvantaged students anywhere in the admissions policy.

Indeed, the section headed Equality and Diversity says:

‘All places at King’s College London Mathematics School are offered on the basis of academic ability and aptitude.’

This does not amount to a commitment to recruit ‘a significant proportion of students from socially disadvantaged backgrounds’, as stated in the headteacher information pack.

A deputy headteacher information pack published in November 2013 had already rowed back from this, simply stating that:

‘Students will be recruited from a wide variety of backgrounds.’

The reasons for such backtracking remain unclear. Perhaps it was only ever insurance against accusations of elitism that never actually materialised.

The website confirms that all students take A levels in maths, further maths and physics, together with an AS EPQ. But now they can also take an optional AS level in computing in Year 12 and may convert it to an A level in Year 13. They will also take either the AEA or STEP papers.

The description of additional curricular provision is somewhat vague. Students will have a series of lessons and educational visits. Each fortnight a KCL lecturer will introduce a new theme, to be explored through ‘mini research projects’. Students will also learn a modern language but to what level is unclear.

A mentor will be assigned to support work for the EPQ. There will also be a maths mentor – always an undergraduate, never ‘a junior member of the maths department’ – available for one meeting a week.

Tuesday afternoons seem to be set aside for sport and exercise. Visual and performing arts will be explored through extra-curricular activity, though this is currently aspirational rather than real:

‘…the school hopes to have sufficient interest to form a student choir, orchestra and dramatic society.’

The length of the school day is six hours and 55 minutes, with five hours of lessons (though the FAQ implies that students will not have a full timetable).

The present staff complement is 10, six of whom seem to be teaching staff. The head was formerly Head of Maths at Highgate School.

Outreach continues for students in Years 10 and 11. There is also a CPD programme for those new to teaching further maths. This is funded by a £75,000 grant from the Mayor’s London Schools Excellence Fund and supports 30 teachers from six schools spread across five boroughs.

KCL’s Access Agreement for 2015/16 says:

‘King’s College London Mathematics School aims to increase substantially the number of young people with the right levels of mathematical attainment to study STEM subjects at top-rated universities. It also aims to improve access to high quality mathematical education at sixth form level and is targeting individuals from schools where such provision is not easily available (in particular, 11-16 schools and schools where further mathematics is not offered as part of the curriculum at A-level). The school has implemented an extensive outreach programme for pupils at KS4, aged 14-16, whereby pupils come to King’s College London for two hours per fortnight over a two-year period. Through this programme, the school will provide students with limited access [sic] to high quality sixth form provision the understanding and skills they need to prepare for A-levels in Maths and Further Maths should they decide to study them, and also to support applications to the maths school should they wish to make them.

The school has also just launched a programme of continuing professional development for maths teachers in London schools. The programme will run for two consecutive years, and will enable high-quality teaching of Further Maths for those new to teaching this A-level. One of the key aims of this programme is to improve take up and retention rates in A-level Further Maths, with a view to increasing numbers of well-trained applicants to STEM subjects at university.’

Exeter

The Exeter School also opened in September 2014, with 34 students, against a planned admission number of 30. Disappointingly only seven are girls. Eleven (32%) are boarders. We do not know the number of applicants.

The School is located in Rougemont House, a Grade 2 listed building close to the University and College. The cost of refurbishment is as yet unknown.

There were relatively fewer details available of Exeter’s plans at the time I wrote my previous post. The January 2013 revealed that:

  • As we have seen, the roll would be 120 students, 60 per year group, with boarding places available for 20% of them.
  • All students would take maths A level and the STEP paper and all would have 1:1 maths mentoring.
  • University academics would provide an ‘enrichment and critical thinking programme’.
  • The Met Office would be involved.

The 2014 admissions policy dates from September 2013.  It indicates that the School will admit 30 students in September 2014, 50 in September 2015 and 60 in September 2016. It will not reach full capacity until September 2017.

Minimum entry requirements are:

  • A* in GCSE Mathematics
  • A or A* in double sciences or single science Physics (in 2015 computer science is also acceptable as an alternative)
  • At least 6 GCSEs at C grade or above, normally to include English Language at a grade B.

So Exeter is more demanding than KCL in respect of the grades required for both GCSE maths and English language, but the minimum number of GCSEs required is one fewer.

The policy says that the School will aim for allocated places to reflect the incidence of potential students across Devon (47%) and in the other three counties served by the school (Cornwall 23%, Somerset 23%, Dorset 6%) but they will not be selected on this basis. There is nothing in the admissions criteria to secure this outcome, so the purpose of this paragraph is unclear.

The selection process involves a written application, a reference an interview and ‘a mathematics-based entry exam’, subsequently called an aptitude test. This is described in identical terms to the test used by KCL – indeed the specimen questions are identical.

The oversubscription criteria involve giving priority to ‘interview answers and the candidates’ potential to thrive and succeed on the course’.

Under ‘Equality and Diversity’ the document says:

‘EMS is committed to widening participation and broadening access to high quality mathematics education. As such, we will target our recruitment in areas which have high levels of deprivation and in schools for which provision is currently limited, such as those without 6th forms.

EMS will encourage applications from female students through targeted marketing and recruitment. However, there will be no positive discrimination for girls in the admissions criteria.’

The first statement is largely meaningless since neither residence in a deprived area nor attendance at a school without a sixth form is mentioned explicitly in the admissions criteria.

The second statement is reflected in the fact that only 20% of the inaugural cohort is female.

The document notes that boarding will be available for learners living more than an hour distant. The proportion of boarders in the first cohort is significantly higher than expected.

It adds that boarding fees will be payable (and published on the School’s website) but it is expected they ‘will be subsidised by a government grant and a private investor’. There will also be a limited number of means-tested full bursaries, the criteria for which will also be published.

At the time of writing neither fees nor subsidies nor bursary criteria are published on the open pages of the website. It also mentions a subsidised transport scheme but provides no details. This is unhelpful to prospective candidates.

Students take A levels in maths and further maths, plus an A level in either physics or computer science. They are also prepared for STEP papers. All students pursue one further AS level at Exeter College, selecting from a choice of over 30 subjects, with the option to complete the A level in Year 13. Amongst the 30 are several non-traditional options such as fashion and design, media studies and world development. The School is clearly not wedded to facilitating subjects!

In maths students will:

‘…collaborate with those in other mathematics schools and meet, converse and work with staff and students from Exeter University’s mathematics department. They will have access to mathematical mentors from the University who will provide 1:1 and small group support for individual development and project work.’

Maths mentors will be 3rd or 4th year undergraduates and sessions will take place fortnightly.

All students will have a pastoral tutor who will ‘deliver a curriculum designed to meet the students’ development needs’. Some extra-curricular options may also be available:

‘Several clubs and societies will exist within EMS, these will be established as a result of students’ own interests. In addition, Exeter College’s specialist facilities, learning centres and other services will be accessible to them. Students will join their friends and other students from the College for sporting and enrichment activities including, for example, structured voluntary work, theatre productions and the Duke of Edinburgh’s Award Scheme.’

I could find no reference to a University-provided enrichment and critical thinking programme or to Met Office involvement.

The Head of Exeter School was formerly a maths teacher and maths AST at Torquay Boys’ Grammar School. Other staff responsibilities are not enumerated, but the Contacts page mentions only one teacher apart from the Head.

Another section of the site says the School will be advertising for a Deputy and ‘teachers of Mathematics, Computer Science and Physics (p/t)’. So the original intention to deploy Exeter College staff seems to have been set aside. Advertisements have been placed for several posts including a Pastoral Leader and an Outreach and Admissions Officer.

An outreach programme is being launched and business links will be established, but there are no details as yet. There are links to a KS4/5 maths teachers’ network sponsored by the Further Maths Support Programme.

Exeter’s 2015/16 Access Agreement says:

‘The University and the College are already joint sponsors of the innovative new Exeter Maths School and are developing a strategic approach to outreach that supports both curriculum enhancement in local schools and progression for the students enrolled in the school. Together with the South Devon UTC, these two new education providers offer opportunities for innovative collaborative approaches to outreach in the region.’

This sounds very much a work in progress.

 

 

Comparing the two schools

My 2013 post observed:

‘From the information so far published, the Exeter project seems very close conceptually to the one at King’s, indeed almost a clone. It would have been good to have seen evidence of a fundamentally different approach.’

If anything, the two projects have grown even more similar as they have matured. To the extent that these are pilot institutions testing out a diversity of models this is not entirely helpful.

Both schools are very small and KCL in particular offers a very restricted range of post-16 qualifications. There is downside to post-16 education on this model – otherwise we wouldn’t be exercised about the negative effects of small sixth forms – though both projects make some effort to broaden their students’ experience and, as we have seen, Exeter includes some shared provision with Exeter College.

The admissions requirements and processes are almost identical. It is important to recognise that neither institution is highly selective, especially in terms of overall GCSE performance and, in this respect, the comparisons with Kolmogorov and other institutions elsewhere in the world are rather misleading.

This is not the top 2% that Cummings cited as the beneficiaries in his essay. Even in terms of mathematical ability, the intake to these schools will be relatively broad.

The expectation that all will take STEP papers may be realistic but, despite the use of an aptitude test, any expectation of universal success is surely over-optimistic.

For Cambridge says STEP papers are ‘aimed at the top 5% or so of all A-level mathematics candidates’.  Fewer than 1,500 students took the most popular Paper 1 in 2013 and, in 2014, over 20% of participants received an Unclassified grade.

Cummings queries my conclusions here

and I have to admit that these are inferred from the evidence set out above. But, on the basis of that evidence, I would be surprised indeed if STEP results for these two schools exceed the national profile in 2016.

Cummings notes that approximately one third of those entered for STEP attend independent schools, meaning that roughly 1,000 of the 2013 cohort were in maintained institutions. There may be some marginal increase in state-funded STEP entry through these two schools, but the impact of MEI support elsewhere is likely to be more significant.

 

The priority attached to excellence is less pronounced than expected. But this is not matched (and justified) by a correspondingly stronger emphasis on equity.

Neither school gives priority within its admissions or oversubscription criteria to students from disadvantaged backgrounds. A major opportunity has been lost as a consequence.

Cummings responds

There are questions to be asked here about just how tightly the universities were held to the specifications they agreed.

There is nothing about the admission of disadvantaged students in the KCL funding agreement (I can’t find Exeter’s). It would be interesting to know what exactly they set down in their proposals, as approved.

One suspects that some effort has been made to prioritise admissions from state schools, especially state schools without sixth forms, but all this is swept up into the interview scores: there is nothing explicit and binding. The fact that 15% of the KCL intake has come from the independent sector shows that  this is insufficient.

Comparison with the admissions policy for the Harris Westminster Sixth Form is instructive:

‘Applicants who have achieved the qualifying score will then be awarded points as follows:

  • One point for the applicant’s home address…if it is in an area of high deprivation, based on an independently published assessment of levels of deprivation of postcodes;
  • One point if they qualify for, or have previously qualified for, Free School Meals.

If Year 12 is oversubscribed then, after the admission of pupils with Special Educational Needs where the Harris Westminster Sixth Form is named on the statement, the criteria will be applied in the order in which they are set out below to those who have achieved a qualifying score:

a. Looked after and former looked after young people;

b. Applicants who have 2 points in accordance with the paragraph above;

c. Applicants who have 1 point in accordance with the paragraph above…’

The funding allocations for academic year 2014/15 show that both maths free schools have been awarded zero free meals funding, suggesting that no pupils eligible for free school meals in Year 11 have been admitted.

So there is too little emphasis on excellence and equity alike. These institutions exemplify a compromise position which, while tenable, will reduce their overall impact on the system.

The only substantive difference between the two schools is that one is located in London and the other in a much more sparsely populated and geographically dispersed region. These latter conditions necessitate a boarding option for some students. The costs associated with boarding are not transparent, but one suspects that they will also serve as a brake on the recruitment of disadvantaged students.

Exeter has no real competitors in its region, other than existing sixth forms and post-16 institutions, but KCL faces stiff competition from the likes of the London Academy of Excellence and the Harris Westminster Sixth Form, both of which are much more substantial institutions offering a wider range of qualifications and, quite possibly, a richer learning experience.

Both Schools are designed to suit students who wish to specialise early and who are content with only limited opportunities to work outside that specialisation. That subgroup does not necessarily include the strongest mathematicians.

It might have been different story if the Schools could have guaranteed progression into the most selective higher education courses, but this they cannot offer. There is no guaranteed progression even to the host universities (whose mathematics departments are not the strongest – one obvious reason why they were attracted to hosting maths schools in the first place).

Exeter and Kings no doubt expect that their Schools will help them to compete more effectively for prospective students – both through direct recruitment and, more indirectly, by raising their profile in the maths education sector – but they will not state this overtly, preferring to emphasise their contribution to improving standards system-wide.

There is no reference to independent evaluation, so one assumes that success indicators will focus on recruitment, a strong showing in the Performance Tables and especially Ofsted inspection outcomes.

A level performance must be consistently high and HE destinations must be commensurate. Because recruitment of disadvantaged students has not been a priority fair access measures are largely irrelevant.

Other indicators should reflect the Schools’ contribution to strengthening the maths talent pipeline and maths education more generally, particularly by offering leadership at regional and national levels.

At this early stage, my judgement is that the KCL project seems rather better placed than Exeter to achieve success. It has hit the ground running while Exeter has some rapid catching up to do. One is good; the other requires improvement.

 

Future Prospects

 

Prospects for the maths school programme

With just seven months before Election Purdah, there is no prospect whatsoever that the programme will reach its target of 12 schools. Indeed it seems highly unlikely that any further projects can be brought to fruition before the end of the spending round, with the possible exception of the mysterious ‘4th proposal’.

On assumes that the Regional Schools Commissioners are now responsible for stimulating and supporting new maths school projects – though this has not been made explicit – but they already have their hands full with many other more pressing priorities.

If Labour were to win the Election it seems unlikely that they would want to extend the programme beyond the schools already established.

Even under the Conservatives it would be extremely vulnerable given its poor track record, the very tight budgetary constraints in the next spending round (especially if schools funding is no longer ringfenced) and the fact that its original champions are no longer in place at DfE.

Cummings suggest that a further five schools might be reasonable objective for the next Parliament, but only if the commitment within DfE is sustained.

Even that unlikely prospect would result in a network of only eight schools by 2020, four short of the original target that was to have been delivered five years earlier.

With the benefit of hindsight one might have taken a different approach to programme design and targeting.  Paradoxically, the centre has appeared overly prescriptive – favouring a ‘Kolmogorov-lite’ model, ideally hosted by a Russell Group institution – but also too vague – omitting to clarify their expectations in a specification with explicit ‘non-negotiables’.

Universities were hesitant to come forward. Some will have had other fish to fry, some may have had reservations arising from fear of elitism, but more still are likely to have been unclear about the Government’s agenda and how best to satisfy it.

The belated decision to flag up the potential contribution to fair access was locking the door after the horse had bolted. Other universities will have noted that neither KCL nor Exeter paid lip service in this direction.

Cummings rejects this analysis. For him the resistance from vice chancellors had a straightforward explanation

According to his narrative, many university mathematicians were on the side of the angels, understanding the advantage to their departments of securing a bigger flow of undergraduates, far better prepared for university study.

But they were thwarted by the corporate centres in their institutions, the vice chancellors hamstrung by their fear of potential reputational damage, invariably associated with the charge of elitism.

Yet I have seen negligible evidence of media criticism of KCL and Exeter on these grounds, or any others for that matter.

The only occasion on which I have seen the term ‘elitism’ wielded is by the massed ranks of the Devon Branch of the NUT. Neither KCL nor Exeter has had to play the trump card of priority for disadvantaged students – indeed I have shown above how they have apparently rowed back from earlier commitments on this front.

We shall probably never know the truth since there are no records of these discussions – and I very much doubt whether any vice chancellors will read this and decide to put the record straight.

My own personal experience has been that, by and large, universities are reluctant to serve as instruments to further government education policy. Their knee-jerk is more of the ‘not invented here’ variety and, even if they are given carte blanche, they remain highly suspicious of government motives. Fundamentally, there is an absence of trust.

An internal champion, such as Alison Wolf at KCL, can help to break this down.

 

There were also policy design issues. Because they were awarded a substantial capital budget – and were wedded to the value of free schools – ministers were driven to focus on creating new stand-alone institutions that might ultimately form a network, rather than on building the network itself.

The decision to create a set of maths hubs was the most sensible place to start, enabling new maths schools to take on the role of hubs when they were ready to do so. But, the maths hubs were a later invention and, to date at least, there have been no efforts to ‘retro-fit’ the maths schools into the network, meaning that these parallel policy strands are not yet integrated.

 

Prospects for the national maths talent pipeline

England is far from having a coherent national strategy to improve maths education or, as one element within that, a convincing plan to strengthen the maths talent pipeline.

Maths education enjoys a surfeit of players with overlapping remits. National organisations include:

A host of other organisations are involved, including the Joint Mathematical Council (JMC), an umbrella body, the Advisory Committee on Mathematics Education (ACME), the United Kingdom Mathematics Trust (UKMT) and the School Mathematics Project (SMP).

This leaves to one side the maths-related element of broader programmes to support between-school collaboration, recruit teachers and develop new-style qualifications. There is a parallel set of equally complex relationships in science education.

Not to put too fine a point on it, there are too many cooks. No single body is in charge; none has lead responsibility for developing the talent pipeline.

Ministers have been energetic in generating a series of stand-alone initiatives. The overarching vision has been sketched out in a series of set-piece speeches, but there is no plan showing how the different elements knit together to create a whole greater than the sum of its parts.

This probably has something to do with an ideological distaste for national strategies of any kind.

The recent introduction of maths hubs might have been intended to bring some much-needed clarity to a complex set of relationships at local, regional and national levels. But the hubs seem to be adding to the complexity by running even more new projects, starting with a Shanghai Teacher Exchange Programme.

Cummings has me down as a hopeless idealist

and who am I to contest his more recent and much more wide-ranging experience? I will only say that I can still recollect the conditions under which many such obstacles can be overcome.

 

courtesy of Jim2K

courtesy of Jim2K

 

Last words

A network-driven approach to talent development might just work – I suggested as much at the end of my previous post – but it must be designed to deliver a set of shared strategic objectives. Someone authoritative needs to hold the ring.

What a pity there wasn’t a mechanism to vire the £72m capital budget for 12 free schools into a pot devoted to this end. For, as things stand, it seems that up to £12m will have been spent on two institutions with a combined annual cohort of 120 students, while as much as £60m may have to be surrendered back to the Treasury.

We are better off than we would have been without the KCL and Exeter Schools, but two schools – or perhaps three – is a drop in the ocean. Even 12 schools of this size would have been hard-pressed to drive improvement across the system.

The failure to capitalise on the potential of these projects to support progression by genuinely disadvantaged students is disappointing and deserves to be revisited.

This might have been a once-in-a-generation chance to mend the maths talent pipeline. I do hope we haven’t blown it.

 

GP

October 2014

16-19 Maths Free Schools Revisited

This post marks the opening of two university-sponsored 16-19 maths free schools with a fresh look at the wider programme that spawned them.

It scrutinises developments since the publication of ‘A Progress Report on 16-19 Maths Free Schools’ (March 2013), building on the foundations within ‘The Introduction in England of Selective 16-19 Maths Free Schools’ (November 2011).

courtesy of Jim2K

courtesy of Jim2K

The broad structure of the post is as follows:

  • A description of the genesis of the programme and a summary of developments up to March 2013.
  • The subsequent history of the programme, from March 2013 to the present day. This reviews efforts to recruit more university sponsors into the programme – and to resist the publication of information showing which had submitted expressions of interest and, subsequently, formal proposals.
  • An assessment of the prospects for the programme at this point and for wider efforts to expand and remodel England’s national maths talent pipeline.

Since many readers will be interested in some of these sections but not others, I have included direct links to the main text from the first word of each bullet point above.

 

Genesis and early developments

Capital investment to support the programme was confirmed in the 2011 Autumn Statement, which referred to:

‘…an extra £600 million to fund 100 additional Free Schools by the end of this parliament. This will include new specialist maths Free Schools for 16-18 year olds, supported by strong university maths departments and academics’.

This followed an orchestrated sequence of stories fed to the media immediately prior to the Statement.

One source reported a plan to establish 12 such schools in major cities by the end of the Parliament (Spring 2015) ‘before the model is expanded nationwide’. These would:

‘…act as a model for similar institutions specialising in other subjects’.

Another confirmed the number of institutions, adding that there would be ‘…a special application process outside the regular free school application process…’

A third added that the project was viewed as an important part of the Government’s strategy for economic growth, suggesting that some of the schools:

‘…would offer pure maths, while others would combine the subject with physics, chemistry or computer sciences.’

Assuming provision for 12 schools at £6m a time, the Treasury had provided a capital budget of £72m available until 2015. It remains unclear whether this sum was ringfenced for university-sponsored maths schools or could be diverted into the wider free schools programme.

We now know that former political adviser Dominic Cummings was a prime instigator of the maths free schools project – and presumably behind the press briefings outlined above.

The most recent edition of his essay ‘Some thoughts on education and political priorities’ (2013) says:

‘We know that at the top end of the ability range, specialist schools, such as the famous Russian ‘Kolmogorov schools’…show that it is possible to educate the most able and interested pupils to an extremely high level…We should give this ~2% a specialist education as per Eton or Kolmogorov, including deep problem-solving skills in maths and physics.

The first English specialist maths schools, run by King’s College and Exeter University, have been approved by the Department for Education and will open in 2014. All of the pupils will be prepared for the maths ‘STEP’ paper that Cambridge requires for entry (or Oxford’s equivalent) – an exam that sets challenging problems involving unfamiliar ways of considering familiar  material, rather than the formulaic multi-step questions of A Level.’

Back in February 2012, TES reported that:

‘The DfE has hosted a consultation meeting on the new free schools with interested parties from the mathematical community in order to outline its plans.’

‘TES understands that officials within the Department for Education are now keen to establish the schools on the model of Kolmogorov, a boarding school that selects the brightest mathematicians in Russia.’

In fact, the meeting discussed a variety of international models and, on 20 February, Education Minister Nick Gibb answered a PQ thus:

‘Alex Cunningham: To ask the Secretary of State for Education when he expects the first free school specialising in mathematics for 16 to 18 year-olds to open; how many 16 to 18 year-olds he expects to enrol in free schools specialising in mathematics by 2015; with which universities he has discussed these free schools; and what guidance he plans to provide to people who wish to apply to open such a school.

Mr Gibb: We are developing proposals on how specialist maths schools for 16 to 18-year-olds might operate and will announce further details in due course. We are keen to engage with all those who have an interest to explore possible models and innovative ideas.’ (Col. 723W).

However, no proposals were published.

The minutes from King’s College London (KCL) Council meeting of 26 June 2012 reveal that:

‘Following approval by the Principal’s Central Team, the College was pursuing discussions with the Department for Education about sponsoring one of 12 specialist Maths schools for 16-18 year olds to be established with the support of university Mathematics departments. The initiative was intended to address national deficiencies in the subject and to promote a flow of highly talented students into university. In discussion, members noted that while the financial and reputational risks and the costs in management time needed to be carefully analysed, the project supported the College’s commitment to widening participation and had the potential to enhance the strengths of the Mathematics Department and the Department of Education and Professional Services, as well as addressing a national problem. The Council approved the College’s continued engagement with this initiative.’

By December 2012 KCL had announced that it would establish a maths free school, with both its maths and education departments involved. The school was scheduled to open in September 2014.

KCL confirmed that it had received from DfE a development grant plus a parallel outreach grant to support a programme for mathematically talented 14-16 year-olds, some of whom might subsequently attend the school.

The minutes of the University of Exeter Council meeting of 13 December 2012 record that:

‘As Council were aware, Exeter was going to be a partner in an exciting regional development to set up one of the first two Maths specialist schools with Exeter College. The other school would be led by King’s College London. This would cater for talented Maths students as a Free School with intake from four counties (Devon, Cornwall, Somerset and Dorset) with a planned total number of students of 120 by September 2017. The bid was submitted to the Department of Education on 11th December and the outcome would be announced in early January, with the school opening in 2014. It would be taught by Exeter College teachers with contributions from staff in pure and applied Maths in the College of Engineering, Mathematics and Physical Sciences (CEMPS), input from the Graduate School of Education and from CEMPS students as mentors and ambassadors. It was hoped that at least some of these talented students would choose to progress to the University. Council would be kept informed of the progress of the bid.’

In January 2013 a DfE press release announced approval of this second school. It would indeed have capacity for 120 students, with Monday-Thursday boarding provision for 20% (24 students), enabling it to recruit from across the four counties named above, so acting as a ‘regional centre of excellence’.

This project had also received a development grant – which we know was up to £300K – had agreement in principle to an outreach grant and also expected to open in September 2014.

There is also reference to plans for Met Office involvement with the School.

The press release repeats that:

‘The ultimate aim is to create a network of schools that operate across England which identify and nurture mathematical and scientific talent.’

A page added to DfE’s website in March 2013 invites further expressions of interest to open maths free schools in September 2014 and beyond.

Parallel Q and A, which has now been removed, made clear that development grants would not be available to new applicants:

‘Is there financial support available to develop our plans?

Not at the beginning. Once we have approved a proposal, we do offer some support to cover the costs of project management, and recruiting some staff before the school opens, in the same way we would for any Free School.’

This has subsequently been reversed (see below).

 

Progress since March 2013

 

The Hard Sell

While KCL and Exeter developed their plans, strenuous efforts were made to encourage other universities to participate in the programme.

A TES piece from May 2013, profiling the newly-appointed head of the KCL school, includes a quote from Alison Wolf – the prominent chair of the project group at KCL:

‘’The Brit School is a really good comparison,” she says. “When we were working on the new school and thinking about what to do, we’d look at their website.

“Maths is very glamorous if you’re a young mathematician, which is why they’ll do well when they are around other people who adore maths.”

The story adds that 16 schools are now planned rather than the original 12, but no source is attributed to this statement.

It seems that the wider strategy at this stage was to convince other potential university sponsors that maths schools were an opportunity not to be missed, to imply that there was already substantial interest from prominent competitors, so encouraging them to climb on board for fear of missing the boat.

 

Playing the Fair Access Card

But there was soon a change of tack. In June 2013, the Guardian reported that education minister Liz Truss had written to the heads of university maths departments to encourage bids.

‘As an incentive to open the new schools, universities will be allowed to fund them using budgets otherwise reserved for improving access to higher education for under-represented and disadvantaged groups….

Les Ebdon, director of Offa, said: “I’d be happy to see more university-led maths free schools because of the role they can play in helping able students from disadvantaged backgrounds access higher education.

“It is for individual universities and colleges to decide whether or not this is something they want to do, but Offa is supportive of anything that is targeted at under-represented groups and helps them to fulfil their potential.”

…According to Truss’s letter, Ebdon confirmed it would be “perfectly legitimate to allocate funding ringfenced for improving access for under-represented groups towards the establishment of such schools,” counting the spending as “widening access”.’

My initial post had pointed to the potential significance of this coupling of excellence and equity as early as November 2011:

‘It is not clear whether a fundamental purpose of these institutions is to support the Government’s drive towards greater social mobility through fair access to competitive universities. However, one might reasonably suggest it would be an oversight not to deploy them…encouraging institutions to give priority during the admissions process would be the likely solution.’

But Ministers’ rather belated conversion to the merits of alignment with social mobility and fair access might have been interpreted as opportunism rather than a sincere effort to join together two parallel strands of Government policy, especially since it had not been identified as a central feature in either KCL’s or Exeter’s plans.

I can find nothing on Offa’s website confirming the statement that funding ringfenced for fair access might be allocated by universities to the development of maths free schools. There is no contemporary press notice and nothing in subsequent guidance on the content of access agreements. This begs the question whether Ebdon’s comments constitute official Offa advice.

However the text of the letter is preserved online and the identical text appears within it:

‘I want to encourage other universities to consider whether they could run similar schools: selective, innovative and stretching our brightest and best young mathematicians. It is a logical extension of the role that dozens of universities have already played in sponsoring academies.

I also wanted to highlight to your colleagues that Professor Les Ebdon, Director of the Office for Fair Access, is enthusiastic about the role university led Maths Free Schools can have in encouraging more young people to go on to study maths at university, and to reap the benefits that brings. Professor Ebdon has also confirmed to me that he considers the sponsorship and development of Maths Free Schools as contributing to higher education ‘widening access’ activity, and that it would be perfectly legitimate to allocate funding ring-fenced for improving access for underrepresented groups towards the establishment of such schools.

Unlike our usual practice for Free Schools, there is no competitive application process for Maths Free Schools. Instead we ask interested universities to submit a short proposal setting out the key features of the school. These proposals need not be long: King’s and Exeter both submitted initial proposals that were around 12 pages…

[There follows a list of bullet points describing the content of these initial proposals, none of which address the admission of students from disadvantaged backgrounds.]

….Both King’s College and the University of Exeter had a number of detailed discussions with colleagues in the Department to develop and refine their proposals and we are always happy to work with universities to help them focus their plans before submitting a formal proposal. If we approve a proposal, we do then offer financial support to cover the costs of project management, and of recruiting some staff before the school opens, in the same we would for any free school.’

(By way of an aside, note that the final emboldened sentence in the quotation above corrects the statement in the Q and A mentioned above. It seems that maths free schools are now treated comparably with all other free school projects in this respect, even though the application process remains different.

The latest version of free school pre-opening guidance gives the sum available in Project Development Grant for 16-19 free schools as £0.25m.)

Going back to Offa, there are no conditions imposed by Ebdon in respect of admissions to the schools, which seems a little over-relaxed, given that they might well attract a predominantly advantaged intake. I wonder whether Ebdon was content to offer personal support but refused to provide official Offa endorsement.

 

 

In July 2013 the BBC reported a speech by Truss at the 2013 ACME Conference. Oddly, the speech is not preserved on the gov.uk site. According to the BBC:

“We want this movement to spread still further,” she told delegates.

“So we’re allowing universities to apply to sponsor new maths free schools through a fast-track, simplified procedure, without having to go through the normal competitive application process.

“These schools will not only improve standards in maths teaching, but will equip talented young people from low-income backgrounds with the skills they need to study maths at university.”

Mrs Truss said the Office for Fair Access had confirmed that, when universities contributed to the sponsorship or development of maths free schools, this would be considered as one of their activities to widen access to under-represented groups – and therefore as part of their access agreement.

“I hope that this is the start of a new network of world-class free schools, under the aegis of top universities, helping to prepare talented 16- to 19-year-olds from any and every background for the demands of university study.”

Note that Ebdon’s endorsement is now Offa’s.

Cummings’ essay remarks in a footnote:

‘Other maths departments were enthusiastic about the idea but Vice Chancellor offices were hostile because of the political fear of accusations of ‘elitism’. Hopefully the recent support of Les Ebdon for the idea will change this.’

A year on, we have no evidence that it has done so.

 

The Soft Sell

By the beginning of the following academic year, a more subtle strategy was adopted. The two schools-in-development launched a maths competition for teams from London and the South-West with prizes awarded by education ministers.

 

 

A November 2013 DfE press release marks the ceremony. Michael Gove is quoted:

‘We need specialist maths free schools like King’s College London (KCL) Maths School and Exeter Mathematics School. They will develop the talents of exceptional young mathematicians and ensure they can compete in the global race.’

The release continues:

‘The KCL and Exeter schools are the first to take advantage of a development grant made available by the Department for Education for the creation of university-led specialist maths free schools.’

The notes include a link to the 1 March webpage mentioned above for ‘Universities interested in developing their own maths free school’.

 

Publicity avoided

We now know that a Freedom of Information request had been submitted to DfE in October 2013, asking how many expressions of interest and firm proposals had been received, which institutions had submitted these and which proposals had been approved and rejected.

The source is an ICO Decision Notice published on 12 June 2014.

The request was initially rejected and this decision was upheld in January 2014 following an internal review. A complaint was immediately lodged with the Information Commissioner’s Office.

The Decision Notice records the Commissioner’s decision that public interest outweighs the case for withholding the information. Accordingly he directs that it should be released to the complainant within 35 calendar days of the date of the Notice (ie by 17 July 2014).

The Notice contains some interesting snippets:

  • ‘It has been the DfE’s experience that interested Heads of Maths have contacted it for further information before seeking to discuss the idea with their Vice Chancellor.’ There is no process for accepting formal expressions of interest.
  • There are…no fixed criteria against which all proposals are assessed.’
  • ‘The DfE confirmed that the application is and has always been the first formal stage of the maths free schools process and it has already stated publicly that it has received three applications from King’s College London, Exeter University and the University of Central Lancashire.’
  • ‘It [ie DfE] confirmed that funding arrangements were only confirmed for the development of maths free schools in February 2014 and many policy decisions on this issue have been shaped by the specifics of the two schools that are due to open soon. It expects the policy to develop even further as more maths free schools are approved.’
  • ‘The DfE explained that universities are extremely risk adverse when it comes to protecting their reputation and so do not want to be publically named until they have submitted an application. As such, if they are named at an earlier point it may make them pull out altogether and may make universities unwilling to approach the DfE with ideas.’
  • ‘Similarly, the DfE argued that if it were to release the reasons why one of the applications was rejected it would be likely to deter future interest as the university would not want the public criticism of its ideas. Given that the policy is driven by university interest, if all potential groups are deterred the policy will fail and students will not be able to enjoy the potential benefits.’

The Commissioner gave these arguments short shrift, pointing out the benefits of transparency for policy development and the encouragement of more successful applications.

The text does not say so explicitly, but one can imagine the Commissioner thinking  ‘given the low level of interest stimulated to date, you might at least try a more open strategy –what have you got to lose?’

It does seem unlikely that university heads of maths departments would submit speculative expressions of interest without internal clearance. Their approaches were presumably of the informal ‘sounding out’ variety. They would understand the shaky internal politics of failing to consult the corporate centre – not to mention their education faculties

The lack of specific and transparent assessment criteria does appear to have backfired. What guarantees might universities otherwise receive that their proposals would be judged objectively?

One can imagine the questions:

  • Is the scheme open to all universities, Russell Group or otherwise?
  • If not, what criteria must the host university satisfy?
  • What counts as a ‘strong mathematics department?’
  • Can projects be led by university departments of education, or only undertaken jointly (as at KCL)?

Without explicit and consistent answers one can readily understand why many universities would be disinclined to pursue the idea.

The reference to belated confirmation of funding arrangements – as recently as February 2014 – is intriguing. It cannot apply to capital funding, unless that was vired in extremis. Perhaps it relates to the parallel recurrent funding pot or simply the availability of project development grants.

The latter seems unlikely given the statement in the letter to HoDOMS, dated some eight months previously.

One suspects that there might have been internal difficulties in ringfencing sufficient recurrent funding to honour proposals as and when they were received. Some prospective bidders might have baulked on being told that their budget could not be confirmed until a later date.

But the eventual resolution of this issue a little over a year before the end of the spending round would be unlikely to have a significant impact on the number of successful bids, especially if unspent capital funding has to be surrendered by Spring 2015.

 

Recent developments

In July 2014 the TES revealed that it had been the source of this FoI request.

 

 

But the story reveals little new, other than that:

‘Five further expressions of interest have been made but not yet yielded an application’

The sources are not revealed.

David Reynolds opines that:

‘Having a small number of schools doesn’t matter if we can get the knowledge from them around the system. So we need them to be excellent schools and we need to somehow get that knowledge around.’

A DfE statement concludes:

‘We continue to welcome applications and expressions of interest from universities and the first maths free schools, set up by two leading universities, will be opening in September.’

So we know there have been eight expressions of interest, three of them converted into firm proposals.

The receipt of the third proposal, from the University of Central Lancashire (UClan), is said to have been made public, but I can find no record of it in the lists of Wave 1 to 7 free school applications so far published.

There is a reference in UCLAN’s 2013-14 access agreement dated 31 May 2012:

‘The University is currently consulting on the formation of a Maths Free School which would be run alongside its new Engineering Innovation Centre at the Preston Campus.’

Nothing is said about the plans in the access agreements for 2014-15 and 2015-16.

There is one further reference on the New Schools Network site to a:

‘Consultant engaged to carry out a feasibility study re a Maths Free School on behalf of the University of Central Lancashire (UCLan)’.

One assumes that this must be out-of-date, unless UCLan is considering a second bid.

Otherwise, a simple process of elimination tells us that UCLan’s proposal must have been rejected. The reason for this is now presumably known to TES, as are the sources of the five expressions of interest that were not converted into proposals. Why have they not published this information?

Perhaps they are waiting for DfE to place these details on its website but, at the time of writing – almost three months after the Decision Notice issued – it has not been uploaded.

Meanwhile, there are no further maths free school proposals in the most recent Wave 7 information relating to applications received by 9 May 2014.

The deadline for Wave 8 is imminent. That may well be the last on this side of the Election.

A further feature published by the TES in October 2014 throws no fresh light on these matters, though it carries a quote by new Secretary of State Nicky Morgan, interviewed at the KCL School launch:

‘I think that some [universities] are clearly waiting to see how the King’s and Exeter schools go. Clearly there is a huge amount of effort required, but I think King’s will be enormously successful, and I am hoping they will be leading by example.’

That sounds suspiciously like tacit admission that there will be no new proposals before a General Election.

Another opinion, diametrically opposed to David Reynolds’ view, is contributed by the head of the school of education at Nottingham University who is also Deputy Chair of ACME:

‘I’m very supportive of more people doing more maths, but even if you have 12 schools, you are really scratching the surface,” said Andrew Noyes, head of the school of education at Nottingham University and a former maths teacher.

“These kinds of policy experiments are very nice and they’re beneficial for a certain number of young people, but they’re relatively cheap compared with providing high-quality maths education at every stage in every school.”’

So what are the prospects for the success of the KCL and Exeter Schools? The next section reviews the evidence so far in the public domain.

 

The KCL and Exeter Free Schools

 

KCL School

The KCL School opened in September 2014 with 68 students, against a planned admissions number of 60. The most recent TES article says that there were 130 applicants and nearly all of those successful were drawn from state schools.

However, another reliable source – a member of the governing body – says that only 85% (ie 58) are from maintained schools, so the independent sector is actually over-represented.

He adds that:

‘Many are from families where neither parent has attended university’

but that is not necessarily an indicator of disadvantage.

We also know that some 43% (29 students) were female, which is a laudable outcome.

The School is located in Lambeth Walk, some distance from KCL’s main campuses. The capital cost of refurbishing the School was seemingly £5m. It occupies two buildings and the main building is shared with a doctor’s surgery.

My March 2013 post summarised KCL’s plans, as revealed by material on the University’s site at that time, supplemented by the content of an information pack for potential heads which is no longer available online.

I have reproduced the main points below, to provide a baseline against which to judge the finished article.

  • The full roll will be 120, with an annual admission number of 60. Potential applicants must have at least 5 GCSE grades A*-C including A*/A in both maths and physics or maths and dual award science.
  • Other admissions criteria will probably include a school reference, ‘our judgement about how much difference attending the school will make to your future based on a number of factors, including the results from an interview’ and the results of a test of aptitude for problem-solving and mathematical thinking.
  • The headteacher information pack adds that ‘the school will also be committed to recruiting a significant proportion of students from socially disadvantaged backgrounds, and to an outreach programme… to further this objective.’
  • All students will take Maths, Further Maths and Physics A levels. They will be expected to take STEP papers and may take a further AS level (an FAQ suggests this will be an Extended Project). Every student will have a maths mentor, either an undergraduate or ‘a junior member of the maths department’.
  • They will also ‘continue with a broad general curriculum, including other sciences, social science, humanities and languages, and have opportunities for sport and the visual and performing arts.’ Some of this provision will be ‘delivered through existing King’s facilities’. The provisional timetable assumes a 40-hour working week, including independent study.
  • The University maths department ‘will be closely involved in curriculum development’ and academics will have ‘regular timetabled contact’, potentially via masterclasses.
  • There will be strong emphasis on collaboration with partner schools. In the longer term, the school ‘intends to seek independent funding for a larger CPD programme associated with the school’s curriculum and pedagogy, and to offer it to a wide range of  schools and students, using school premises out of hours’.

At the time of writing, the KCL Maths School website does not have a working link to the admissions policy, although it can be found online.

As expected, 60 students will be admitted in September 2015. Minimum requirements are now

‘A or A* in GCSE Mathematics or in iGCSE Mathematics

Either an A or A* in GCSE Physics or iGCSE Physics, or an AA, A*A or A*A* in GCSE Science and GCSE Additional Science, or an A or A* in all three Physics modules contained within the GCSE Science, Additional Science and Further Additional Science qualifications; and

A*-C grade in 5 other GCSEs or other qualifications that count towards the Key Stage 4 performance tables compiled by the Department of Education, normally including English language.’

So the minimum requirement has been stiffened to at least seven GCSEs, or equivalent, including A*/A grades in maths and physics and at least a C in English language.

The application process does indeed include a reference, an aptitude test and an interview.

The test is based on KS3 national curriculum material up to Level 8, containing ‘routine and less familiar problems’. Some specimen questions are supplied.

The latest TES story says there are two interviews but this is wrong – there is one interview but two interview scores. One of the two scores is ‘to assess to what extent the school is likely to add value in terms of making a difference to [candidates’] future careers’ but there is no explicit reference to priority for disadvantaged students anywhere in the admissions policy.

Indeed, the section headed Equality and Diversity says:

‘All places at King’s College London Mathematics School are offered on the basis of academic ability and aptitude.’

This does not amount to a commitment to recruit ‘a significant proportion of students from socially disadvantaged backgrounds’, as stated in the headteacher information pack.

The website confirms that all students take A levels in maths, further maths and physics, together with an AS EPQ. But now they can also take an optional AS level in computing in Year 12 and may convert it to an A level in Year 13. They will also take either the AEA or STEP papers.

The description of additional curricular provision is somewhat vague. Students will have a series of lessons and educational visits. Each fortnight a KCL lecturer will introduce a new theme, to be explored through ‘mini research projects’. Students will also learn a modern language but to what level is unclear.

A mentor will be assigned to support work for the EPQ. There will also be a maths mentor – always an undergraduate, never ‘a junior member of the maths department’ – available for one meeting a week.

Tuesday afternoons seem to be set aside for sport and exercise. Visual and performing arts will be explored through extra-curricular activity, though this is currently aspirational rather than real:

‘…the school hopes to have sufficient interest to form a student choir, orchestra and dramatic society.’

The length of the school day is six hours and 55 minutes, with five hours of lessons (though the FAQ implies that students will not have a full timetable).

The present staff complement is 10, six of whom seem to be teaching staff. The head was formerly Head of Maths at Highgate School.

Outreach continues for students in Years 10 and 11. There is also a CPD programme for those new to teaching further maths. This is funded by a £75,000 grant from the Mayor’s London Schools Excellence Fund and supports 30 teachers from six schools spread across five boroughs.

KCL’s Access Agreement for 2015/16 says:

‘King’s College London Mathematics School aims to increase substantially the number of young people with the right levels of mathematical attainment to study STEM subjects at top-rated universities. It also aims to improve access to high quality mathematical education at sixth form level and is targeting individuals from schools where such provision is not easily available (in particular, 11-16 schools and schools where further mathematics is not offered as part of the curriculum at A-level). The school has implemented an extensive outreach programme for pupils at KS4, aged 14-16, whereby pupils come to King’s College London for two hours per fortnight over a two-year period. Through this programme, the school will provide students with limited access [sic] to high quality sixth form provision the understanding and skills they need to prepare for A-levels in Maths and Further Maths should they decide to study them, and also to support applications to the maths school should they wish to make them.

The school has also just launched a programme of continuing professional development for maths teachers in London schools. The programme will run for two consecutive years, and will enable high-quality teaching of Further Maths for those new to teaching this A-level. One of the key aims of this programme is to improve take up and retention rates in A-level Further Maths, with a view to increasing numbers of well-trained applicants to STEM subjects at university.’

Exeter

The Exeter School also opened in September 2014, with 34 students, against a planned admission number of 30. Disappointingly only seven are girls. Eleven (32%) are boarders. We do not know the number of applicants.

The School is located in Rougemont House, a Grade 2 listed building close to the University and College. The cost of refurbishment is as yet unknown.

There were relatively fewer details available of Exeter’s plans at the time I wrote my previous post. The January 2013 revealed that:

  • As we have seen, the roll would be 120 students, 60 per year group, with boarding places available for 20% of them.
  • All students would take maths A level and the STEP paper and all would have 1:1 maths mentoring.
  • University academics would provide an ‘enrichment and critical thinking programme’.
  • The Met Office would be involved.

The 2014 admissions policy dates from September 2013.  It indicates that the School will admit 30 students in September 2014, 50 in September 2015 and 60 in September 2016. It will not reach full capacity until September 2017.

Minimum entry requirements are:

  • A* in GCSE Mathematics
  • A or A* in double sciences or single science Physics (in 2015 computer science is also acceptable as an alternative)
  • At least 6 GCSEs at C grade or above, normally to include English Language at a grade B.

So Exeter is more demanding than KCL in respect of the grades required for both GCSE maths and English language, but the minimum number of GCSEs required is one fewer.

The policy says that the School will aim for allocated places to reflect the incidence of potential students across Devon (47%) and in the other three counties served by the school (Cornwall 23%, Somerset 23%, Dorset 6%) but they will not be selected on this basis. There is nothing in the admissions criteria to secure this outcome, so the purpose of this paragraph is unclear.

The selection process involves a written application, a reference an interview and ‘a mathematics-based entry exam’, subsequently called an aptitude test. This is described in identical terms to the test used by KCL – indeed the specimen questions are identical.

The oversubscription criteria involve giving priority to ‘interview answers and the candidates’ potential to thrive and succeed on the course’.

Under ‘Equality and Diversity’ the document says:

‘EMS is committed to widening participation and broadening access to high quality mathematics education. As such, we will target our recruitment in areas which have high levels of deprivation and in schools for which provision is currently limited, such as those without 6th forms.

EMS will encourage applications from female students through targeted marketing and recruitment. However, there will be no positive discrimination for girls in the admissions criteria.’

The first statement is largely meaningless since neither residence in a deprived area nor attendance at a school without a sixth form is mentioned explicitly in the admissions criteria.

The second statement is reflected in the fact that only 20% of the inaugural cohort is female.

The document notes that boarding will be available for learners living more than an hour distant. The proportion of boarders in the first cohort is significantly higher than expected.

It adds that boarding fees will be payable (and published on the School’s website) but it is expected they ‘will be subsidised by a government grant and a private investor’. There will also be a limited number of means-tested full bursaries, the criteria for which will also be published.

At the time of writing neither fees nor subsidies nor bursary criteria are published on the open pages of the website. It also mentions a subsidised transport scheme but provides no details. This is unhelpful to prospective candidates.

Students take A levels in maths and further maths, plus an A level in either physics or computer science. They are also prepared for STEP papers. All students pursue one further AS level at Exeter College, selecting from a choice of over 30 subjects, with the option to complete the A level in Year 13. Amongst the 30 are several non-traditional options such as fashion and design, media studies and world development. The School is clearly not wedded to facilitating subjects!

In maths students will:

‘…collaborate with those in other mathematics schools and meet, converse and work with staff and students from Exeter University’s mathematics department. They will have access to mathematical mentors from the University who will provide 1:1 and small group support for individual development and project work.’

Maths mentors will be 3rd or 4th year undergraduates and sessions will take place fortnightly.

All students will have a pastoral tutor who will ‘deliver a curriculum designed to meet the students’ development needs’. Some extra-curricular options may also be available:

‘Several clubs and societies will exist within EMS, these will be established as a result of students’ own interests. In addition, Exeter College’s specialist facilities, learning centres and other services will be accessible to them. Students will join their friends and other students from the College for sporting and enrichment activities including, for example, structured voluntary work, theatre productions and the Duke of Edinburgh’s Award Scheme.’

I could find no reference to a University-provided enrichment and critical thinking programme or to Met Office involvement.

The Head of Exeter School was formerly a maths teacher and maths AST at Torquay Boys’ Grammar School. Other staff responsibilities are not enumerated, but the Contacts page mentions only one teacher apart from the Head.

Another section of the site says the School will be advertising for a Deputy and ‘teachers of Mathematics, Computer Science and Physics (p/t)’. Advertisements have been placed for several posts including a Pastoral Leader and an Outreach and Admissions Officer.

An outreach programme is being launched and business links will be established, but there are no details as yet. There are links to a KS4/5 maths teachers’ network sponsored by the Further Maths Support Programme.

Exeter’s 2015/16 Access Agreement says:

‘The University and the College are already joint sponsors of the innovative new Exeter Maths School and are developing a strategic approach to outreach that supports both curriculum enhancement in local schools and progression for the students enrolled in the school. Together with the South Devon UTC, these two new education providers offer opportunities for innovative collaborative approaches to outreach in the region.’

This sounds very much a work in progress.

 

Comparing the two schools

My 2013 post observed:

‘From the information so far published, the Exeter project seems very close conceptually to the one at King’s, indeed almost a clone. It would have been good to have seen evidence of a fundamentally different approach.’

If anything, the two projects have grown even more similar as they have matured. To the extent that these are pilot institutions testing out a diversity of models this is not entirely helpful.

Both Schools are very small and KCL in particular offers a very restricted range of post-16 qualifications. There is downside to post-16 education on this model – otherwise we wouldn’t be exercised about the negative effects of small sixth forms – though both projects make some effort to broaden their students’ experience and, as we have seen, Exeter includes some shared provision with Exeter College.

The admissions requirements and processes are almost identical. It is important to recognise that neither institution is highly selective, especially in terms of overall GCSE performance and, in this respect, the comparisons with Kolmogorov and other institutions elsewhere in the world are rather misleading.

This is not the top 2% that Cummings cited as the beneficiaries in his essay. Even in terms of mathematical ability, the intake to these schools will be relatively broad.

The expectation that all will take STEP papers may be realistic but, despite the use of an aptitude test, any expectation of universal success is surely over-optimistic.

For Cambridge says STEP papers are ‘aimed at the top 5% or so of all A-level mathematics candidates’.  Fewer than 1,500 students took the most popular Paper 1 in 2013 and, in 2014, over 20% of participants received an Unclassified grade.

Cummings notes that approximately one third of those entered for STEP attend independent schools, meaning that roughly 1,000 of the 2013 cohort were in maintained institutions. There may be some marginal increase in state-funded STEP entry through these two schools, but the impact of MEI support elsewhere is likely to be more significant.

The priority attached to excellence is less pronounced than expected. But this is not matched (and justified) by a correspondingly stronger emphasis on equity.

Neither school gives priority within its admissions or oversubscription criteria to students from disadvantaged backgrounds. A major opportunity has been lost as a consequence.

So there is insufficient emphasis on excellence and equity alike. These institutions exemplify a compromise position which, while tenable, will reduce their overall impact on the system.

The only substantive difference between the two schools is that one is located in London and the other in a much more sparsely populated and geographically dispersed region. These latter conditions necessitate a boarding option for some students. The costs associated with boarding are not transparent, but one suspects that they will also serve as a brake on the recruitment of disadvantaged students.

Exeter has no real competitors in its region, other than existing sixth forms and post-16 institutions, but KCL faces stiff competition from the likes of the London Academy of Excellence and the Harris Westminster Sixth Form, both of which are much more substantial institutions offering a wider range of qualifications and, quite possibly, a richer learning experience.

Both Schools are designed to suit students who wish to specialise early and who are content with only limited opportunities to work outside that specialisation. That subgroup does not necessarily include the strongest mathematicians.

It might have been different story if the Schools could have guaranteed progression into the most selective higher education courses, but this they cannot offer. There is no guaranteed progression even to the host universities (whose mathematics departments are not the strongest – one obvious reason why they were attracted to hosting maths schools in the first place).

Exeter and Kings no doubt expect that their Schools will help them to compete more effectively for prospective students – both through direct recruitment and, more indirectly, by raising their profile in the maths education sector – but they will not state this overtly, preferring to emphasis their contribution to improving standards system-wide.

There is no reference to independent evaluation, so one assumes that success indicators will focus on recruitment, a strong showing in the Performance Tables and especially Ofsted inspection outcomes.

A level performance must be consistently high and HE destinations must be commensurate. Because recruitment of disadvantaged students has not been a priority fair access measures are largely irrelevant.

Other indicators should reflect the Schools’ contribution to strengthening the maths talent pipeline and maths education more generally, particularly by offering leadership at regional and national levels.

At this early stage, my judgement is that the KCL project seems rather better placed than Exeter to achieve success. It has hit the ground running while Exeter has some rapid catching up to do. One is good; the other requires improvement.

 

Future Prospects

 

Prospects for the maths school programme

With just seven months before Election Purdah, there is no prospect whatsoever that the programme will reach its target of 12 schools. Indeed it seems highly unlikely that any further projects can be brought to fruition before the end of the spending round.

On assumes that the Regional Schools Commissioners are now responsible for stimulating and supporting new maths school projects – though this has not been made explicit – but they already have their hands full with many other more pressing priorities.

If Labour were to win the Election it seems unlikely that they would want to extend the programme beyond the two schools already established.

Even under the Conservatives it would be extremely vulnerable given its poor track record, the very tight budgetary constraints in the next spending round (especially if schools funding is no longer ringfenced) and the fact that its original champions are no longer in place at DfE.

With the benefit of hindsight one might have taken a different approach to programme design and targeting.  Paradoxically, the centre has appeared overly prescriptive – favouring a ‘Kolmogorov-lite’ model, ideally hosted by a Russell Group institution – but also too vague – omitting to clarify their expectations in a specification with explicit ‘non-negotiables’.

Universities were hesitant to come forward. Some will have had other fish to fry, some may have had reservations arising from fear of elitism, but more still are likely to have been unclear about the Government’s agenda and how best to satisfy it.

The belated decision to flag up the potential contribution to fair access was locking the door after the horse had bolted. Other universities will have noted that neither KCL nor Exeter paid lip service in this direction.

Because they were awarded a substantial capital budget – and were wedded to the value of free schools – ministers were driven to focus on creating new stand-alone institutions that might ultimately form a network, rather than on building the network itself.

The decision to create a set of maths hubs was the most sensible place to start, enabling new maths schools to take on the role of hubs when they were ready to do so. But, the maths hubs were a later invention and, to date at least, there have been no efforts to ‘retro-fit’ the maths schools into the network, meaning that these parallel policy strands are not yet integrated.

 

Prospects for the national maths talent pipeline

England is far from having a coherent national strategy to improve maths education or, as one element within that, a convincing plan to strengthen the maths talent pipeline.

Maths education enjoys a surfeit of players with overlapping remits. National organisations include:

A host of other organisations are involved, including the Joint Mathematical Council (JMC), an umbrella body, the Advisory Committee on Mathematics Education (ACME), the United Kingdom Mathematics Trust (UKMT) and the School Mathematics Project (SMP).

This leaves to one side the maths-related element of broader programmes to support between-school collaboration, recruit teachers and develop new-style qualifications. There is a parallel set of equally complex relationships in science education.

Not to put to finer point on it, there are too many cooks. No single body is in charge; none has lead responsibility for developing the talent pipeline.

Ministers have been energetic in generating a series of stand-alone initiatives. The overarching vision has been sketched out in a series of set-piece speeches, but there is no plan showing how the different elements knit together to create a whole greater than the sum of its parts.

This probably has something to do with an ideological distaste for national strategies of any kind.

The recent introduction of maths hubs might have been intended to bring some much-needed clarity to a complex set of relationships at local, regional and national levels. But the hubs seem to be adding to the complexity by running even more new projects, starting with a Shanghai Teacher Exchange Programme.

A network-driven approach to talent development might just work – I suggested as much at the end of my previous post – but it must be designed to deliver a set of shared strategic objectives. Someone authoritative needs to hold the ring.

What a pity there wasn’t a mechanism to vire the £72m capital budget for 12 free schools into a pot devoted to this end. For, as things stand, it seems that up to £12m will have been spent on two institutions with a combined annual cohort of 120 students, while a further £60m may have to be surrendered back to the Treasury.

We are better off than we would have been without the KCL and Exeter Schools, but two schools is a drop in the ocean. Even 12 schools of this size would have been hard-pressed to drive improvement across the system.

This might have been a once-in-a-generation chance to mend the maths talent pipeline. I hope we haven’t blown it.

 

GP

October 2014

A Progress Report on 16-19 Maths Free Schools

.

Andrey Kolmogorov courtesy of Svjo

Andrey Kolmogorov courtesy of Svjo

Written on the eve of the 2013 Budget, this post is a progress report on the development of a network of selective 16-19 maths free schools, set in the wider context of the economic arguments for investment in gifted education.

I don’t anticipate a postscript detailing substantive new policy announcements within the Chancellor’s Budget Statement tomorrow. Nor is it likely that further support will be directed towards this existing initiative, given that little of the existing budget has been used up to date.

I set out below the information currently in the public domain and offer a provisional yet constructive assessment of how the 16-19 maths free school project is shaping up.

.

Announcement of 16-19 Maths Free Schools

Back in November 2011 I devoted a post to the announcement of the introduction of a cadre of selective 16-19 maths free schools in England, as outlined in the Chancellor’s 2011 Autumn Statement.

A proportion of a £600m allocation to meet the capital costs of 100 free schools was notionally earmarked for ‘New Maths Free Schools for 16-18 year-olds’ to be ‘supported by strong university maths departments and academics‘.

The announcement suggested these would be:

‘Exactly what Britain needs to match our competitors – and produce more of the engineering and science graduates so important for our longer term economic success.’

Well-informed press reports prior to the announcement suggested that there would be at least 12 schools and the resulting network would serve as a model that might be extended to other subjects.

It was suggested that the first would be located in major cities. Some might focus solely on maths and others on a wider STEM curriculum but they would all prepare students to excel at top universities and in subsequent IT, academic or entrepreneurial careers.

Assuming a network of 12 schools and £6m per school, the capital funding notionally set aside for this purpose amounts to £72 million. This is presumably available until the end of the current spending review cycle, so would have to be allocated by Spring 2015 at the latest.

.

Progress since the Announcement

We are now some 16 months on from the announcement and, with almost exactly two years until the end of the spending review period, we are probably about 40% through the project as currently funded.

So it seemed timely to review progress to date.

A handful of 16-19 free schools with a STEM specialism are in the pipeline – including what is now called the Sir Isaac Newton Sixth Form in Norwich and STEM6 in Camden, London. But these are slightly different animals, falling outside the project under discussion because they are not supported by university maths departments.

We know from a FAQ briefing published by DfE that:

‘The common feature of all specialist maths Free Schools is significant involvement from a university maths department. Universities can apply to set up a specialist maths Free School on their own, or in partnership with another strong education provider. Similar specialist maths schools, with significant input from universities, already operate in the United States, Russia and China.’

The development of these institutions is described as ‘a pathfinder programme’, which explicitly implies that the model may be extended if successful.

Interested universities are invited to submit brief proposals to a specialist support team whose home page says:

‘Maths is a strategic priority in education and is at the heart of improving our society and economy. This country has some brilliant university maths departments and world-famous mathematicians, but they have become disconnected from schools, school curriculums and exams.

The new specialist maths schools aim to bridge the gap between school and university maths, and in doing so, demonstrate how new approaches can bring dramatic improvements in performance that can be applied more widely.’

Applications are invited to open further schools ‘in September 2014 and beyond’.

I say further because a January 2013 press release celebrates the first two successful applications, submitted by King’s College London and the University of Exeter in the South-West.

This tells us that:

‘The ultimate aim is to create a network of schools that operate across England which identify and nurture mathematical and scientific talent. This is similar to the Russian model, which includes the renowned Kolmogorov School in Moscow, established by Andrei Kolmogorov – one of the 20th century’s most respected mathematicians.’

The shift from discussion of a network to a single Russian school is something of a logical non-sequitur, and it is not clear why Kolmogorov is singled out when there are so many alternative models worldwide.

The Kolmogorov theme is further developed in a TES story from February 2012 which reports that:

‘The DfE has hosted a consultation meeting on the new free schools with interested parties from the mathematical community in order to outline its plans.

Professor Alexandre Borovik, an expert on selective maths schools who teaches at the University of Manchester, attended the meeting and was encouraged by the government’s plans.

“So far, it has been only independent schools that have been able to produce mathematicians on anything like a similar scale, but there has been nothing like it in the state sector,” Professor Borovik said. “To see whether it can be done, you really have to be very selective and go down the route of what was successful in Eastern Europe and Russia.”’

The press release also places this initiative in the context of ‘the government’s strategy to increase universities’ involvement in what pupils learn before applying for a university place’ and wider plans ‘to boost maths education’.

The mid-section of this post draws together currently available information about the two live projects.

.

From Kew Gardens courtesy of Gifted Phoenix

From Kew Gardens courtesy of Gifted Phoenix

 .

King’s College London Mathematics School (KCLMS)

 .

Press Release

King’s College (KCL) published a press release on 14 December 2012 confirming that it had received a development grant for its planned school which would open in September 2014.

I note in passing that the Department’s FAQ briefing contains the following Q and A:

Is there financial support available to develop our plans?

Not at the beginning. Once we have approved a proposal, we do offer some support to cover the costs of project management, and recruiting some staff before the school opens, in the same way we would for any Free School.’

which would suggest that the development grants made available for the first two projects are not available to support new proposals.

KCL’s press release suggests that the school will contribute to the Government’s plans:

‘to improve mathematics education in the state sector and increase the number of mathematically talented young people with the right levels of attainment to study STEM subjects at top-rated universities…

…It will aim to cater for students who have both exceptional ability in Mathematics and an intense interest in the subject, and to allow them to study with a critical mass of students with a similar passion for Mathematics.’

There is a quotation from Secretary of State Michael Gove:

‘I am delighted that King’s College London is going to open a specialist maths Free School. If we are to find a future Fields Medallist in our schools, we have to raise standards in maths teaching and create an environment that allows the most gifted to flourish…’

The release explains that:

  • The Project involves KCL’s Department of Educational and Professional Studies as well as its Department of Mathematics and is led by Alison Wolf, Professor of Public Sector Management, perhaps best known as author of the Wolf Review of Vocational Education, commissioned by the Government shortly after it came to power.
  • KCL has also been awarded ‘an outreach grant’ by DfE ‘to support work with mathematically talented 14-16 year-olds in schools without high levels of specialist Mathematics teaching’. This builds on an existing programme called The King’s Factor  targeted at Years 12-13. It implies that the outreach programme will be used to ‘talent spot’ potential candidates and act as a feeder for the new free school.
  • The school is likely to be located close to KCL’s Waterloo Campus ‘a transport hub easily reached from a very large part of the greater London area. The school will therefore be able to draw on a wide catchment area in which there are large numbers of prospective high-attaining students.’

.

Web Pages

KCL’s website also has a set of pages devoted to the new School which adds a few additional snippets of information.

It says the University announced the new school on 14 November, a month before the date of the press release. I think this must be an error.

The ‘initial setting up’ is being undertaken by KCL’s senior management team (which does not seem to contain Alison Wolf, previously named as the project lead).

Several potential sites in and around Waterloo are under consideration by KCL and DfE.

(The location and securing of suitable sites has been a particular problem for new free schools in London, though the final report of the Mayor’s Education Inquiry proposed action to address this.)

All students attending the School will take the same three A levels – Maths, Further Maths and Physics. They will also be expected to take STEP papers and ‘may take another AS level’ (The range of available options is not specified, but a subsequent FAQ section suggests the choice will probably be confined to the Extended Project.)

Otherwise students:

‘will continue with a broad general curriculum, including other sciences, social science, humanities and languages, and have opportunities for sport and the visual and performing arts. Some of these subjects will be delivered through existing King’s facilities.

Through this broader curriculum and learning to see the world through different disciplinary perspectives, the school will foster intellectual curiosity, clear and independent thought, creativity and a sense of social responsibility.’

Exactly how these additional elements will be fitted into the timetable is not explained.

The school roll will be 120 students – 60 per year. In the first year of operation there will be only one intake, so full complement will not be reached until AY2015/16.

The KS4 outreach programme began in September 2012, so has a full two years of operation before the School opens, enabling it to pick up promising candidates at the start of Year 10.

It is:

‘designed to have a positive effect on the people involved, even if they do not wish to apply to KCLMS or are unsuccessful in the selection process.’ [my emphasis]

The FAQ makes clear that graduates of the school will not necessarily be expected to continue their undergraduate studies at King’s (though the project is clearly attractive precisely because it should help to provide them with a richer pool of applicants).

There is no suggestion that graduates of the School will have preferred status in admission to the University (though that might have been an option, especially for those from disadvantaged backgrounds).

It is also clear that the School will not be suitable for intending medical students:

‘In the main, we expect students to go on to study Maths, Physics, Engineering, Statistics or Computer Science.’

Potential students are invited to apply online from 30 September 2013.  They must have at least 5 GCSE grades A*-C including A*/A in both maths and physics or maths and dual award science. Oddly, GCSE English is not a requirement but ‘will normally be one of those grades’.

These are not particularly demanding requirements, potentially hard to reconcile with the reference to ‘exceptional ability’ above and the comparison with Kolmogorov. Further comment on the pitch of these selection criteria is provided below.

Other admissions criteria are not finalised but will probably include a school reference, ‘our judgement about how much difference attending the school will make to your future based on a number of factors, including the results from an interview’ and the results of a maths aptitude test that will assess problem-solving and mathematical thinking.

Every student will have a maths mentor, either an undergraduate or ‘a junior member of the maths department’. It is not clear whether this is one-to-one provision.

A headteacher will be appointed in April 2013, to take up post in September 2013 and there will be open evenings for prospective students and their families in October and November.

A ‘latest news’ section contains links to various pieces of media coverage about the School. Some are behind paywalls but those that are accessible repeat the information set out in the press release and summarised above.

.

Headteacher Job Pack

Further details are however available in the job pack for the Headteacher vacancy.

This explains that KCLMS:

‘will be run by a Trust, which the College expects to establish in late March 2013; and by a Board of Governors. This appointment is being managed by King’s College London pending the formal establishment of the trust and the signing of the Funding Agreement between the King’s College London Mathematics School Trust and the Department for Education. The person appointed to this position will be employed by the trust.’

It says that ‘students will be recruited from a wide variety of backgrounds’ adding that:

‘The school will also be committed to recruiting a significant proportion of students from socially disadvantaged backgrounds, and to an outreach programme… to further this objective.’

But this ‘significant proportion’ is not quantified. Unless it is truly significant – perhaps  a third of available places – the School could very easily become monopolised by the ‘sharp-elbowed middle classes’ or even by students transferring from the independent sector.

The curriculum will not be accelerative:

‘The aim will not be to cover A-level mathematics rapidly and then start on first year university material, but to teach mathematics which includes the A-level material in a way which develops mathematical thinking and an understanding of the logical connections within the subject….

Thus the material covered will be close to that in A-level maths, but the style of study will be different to that in most schools. Particular features will be:

  • Much greater mathematical rigour, and a general supposition that statements must be proved and methods justified;
  • An intellectual approach, putting work in mathematical and historical contexts;
  • Applications informed by current use of mathematics;
  • Integration of methods and ideas used in computer science.
  • Examinations being seen as hurdles to be taken in the students’ stride, not high jumps to intimidate and confound.’

More on this below.

The provisional timetable is based on a 40-hour working week, including independent study.

This will not be an autonomous institution – the University will be very much ‘hands on’:

‘The Mathematics department of King’s College London will be closely involved in curriculum development for the school, both before and after opening, ensuring strong intellectual foundations and insight into developing applications of mathematics’

Academics will also have ‘regular timetabled contact’, potentially via masterclasses.

There will be strong emphasis on collaboration ‘with other schools and teachers who are interested in developing new pedagogies.’ In addition to continuing the existing outreach programme, it is intended that there will be further engagement for students and teachers alike.

There is reference to a network of schools that ‘could provide a valuable means of sharing expertise and good practice, supporting the professional development of teachers at KCLMS and elsewhere.’

Moreover:

‘In the longer term, the school intends to seek independent funding for a larger CPD programme associated with the school’s curriculum and pedagogy, and to offer it to a wide range of  schools and students, using school premises out of hours. This will contribute directly to schools’ teaching quality (and results), and is an important direct benefit that can be offered in return for schools’ collaboration in identifying potential students.’

.

From Kew Gardens courtesy of Gifted Phoenix

From Kew Gardens courtesy of Gifted Phoenix

.

Exeter University’s Specialist Maths Free School

There is much less information so far in the public domain about the parallel institution at Exeter.

We know from the University’s press release of 21 January that the project is a partnership between the University and Exeter College, a tertiary institution providing a range of post-16 and higher education courses.

The new institution will be based in Exeter and is also scheduled to open in September 2014.

The number on roll will again be 120 – 60 in each year group, but 20% of places (so approximately 24) will be boarding places, with students staying at the University from Monday to Thursday in term-time. This will enable students from across the region to attend and implies a compacted four-day timetable, perhaps complemented by independent study on Friday’s journey home and over the weekend.

The new School is described as ‘a regional centre of excellence’ supported by the mathematical strength of the University and the College’s ‘curricular and pastoral support’. These partners have also received a development grant to underwrite their project (see comment above about that provision apparently being removed for subsequent proposals).

Few further details are provided, other than that:

  • Students here will also be encouraged to take STEP papers.
  • The University will provide a proportion of the teaching: ‘at least 13 hours of maths, physics and computer science teaching a week’ and ‘students will be exposed to mathematical problem-solving’.
  • The University will also offer:

.‘An enrichment and critical thinking programme. The emphasis will be on applied maths, with students given the opportunity to work with academics to apply mathematical concepts to scientific research on subjects like advanced engineering.

  • Students will also benefit from ‘one-to-one “maths mentoring”’.
  • The Met Office ‘hopes to involve the Free School students in its work’. (The Met Office College is based in Exeter.) This sounds highly provisional.
  • There is also agreement ‘in principle’ from DfE to pay an outreach grant which will ‘support the teaching of maths in schools in the region, running maths workshops and to identify potential applicants’. The University’s existing outreach effort seems fairly limited

DfE’s press release contains identical information and little more is revealed in the wider press coverage.

These plans are obviously still at a very early stage – although there must have been significantly more detail in the papers submitted for DfE approval – and there has been no update since the announcement.

From the information so far published, the Exeter project seems very close conceptually to the one at King’s, indeed almost a clone. It would have been good to have seen evidence of a fundamentally different approach.

We do not know whether the University’s School of Education will be directly involved (though, interestingly – and perhaps tellingly – its news section makes no reference to the free school, preferring to highlight instead an entirely different initiative).

.

The Level and Source of Interest from Universities

Sixteen months on from the announcement, initial confirmation of just two projects – both of them still subject to approval of their funding agreements – is arguably indicative of limited interest from potential host universities, despite the very generous capital and recurrent funding available.

There may be some ideological opposition to free schools in some universities, but that is unlikely to be the principal cause of their apparent hesitancy to come forward.

Part of the problem is that the Government is fishing in a small pool. References to ‘leading university maths departments’ and ‘world class institutions’ is rather transparent code for the Russell Group, an organisation comprising 24 universities, just 20 of them in England.

Ministers have been criticised for focusing their policies exclusively on this subset of universities, on the assumption that membership defines higher education quality, when in practice there are weaknesses in some Russell Group provision and exceptionally strong provision in most if not all universities outside the Group.

Even in maths, some universities outside the Russell Group are placed highly in national rankings.

In this example the top 20 includes the Universities of Bath (7), Lancaster (12), Southampton (16), Surrey (18), Loughborough (19) and Kent (20). None are members of the Russell Group.

In comparison, King’s and Exeter are ranked 22 and 25 respectively.

Moreover, each has regional competitors placed higher up the rankings than they are. In London there is Imperial (6), LSE (8) and UCL (11). In the South-West there is: Bath (7), Bristol (9) and, arguably Southampton (16).

These are not idiosyncratic results. If we apply an alternative ranking, Exeter is placed 17th and King’s 18th. In the South-West, Bath is 7th and Bristol 11th. In London, Imperial is 3rd, LSE 14thand UCL 15th. (Non-Russell Group institutions outranking Exeter and Kings include Bath, Lancaster, UEA, Loughborough and Surrey.)

Both Kings and Exeter are therefore likely to be attracted to this initiative because they anticipate that it will help them in future to secure a relatively larger share of the best students, so enabling them to compete more effectively with their better-placed competitors.

On this evidence, the scheme is most likely to attract other Russell Group institutions with a similar mid-table profile in other regions – maybe the likes of Liverpool (35 and 41), Birmingham (32 and 26), York (30 and 21), Sheffield (29 and 26) and Manchester (26 and 30).

It might help the Government to spell out explicitly that they are not interested solely in Russell Group institutions, recognising that excellent maths provision exists elsewhere. It might also help to offer some explicit guidance on the thresholds that they expect such maths departments to exceed.

The trouble is that there is a bewildering array of alternative models already being pursued by universities:

  • Many are involved in the development of a subset of University Technical Schools (UTCs) – the current list of projects is also available from this link.
  • A few are interested in another project which has so far attracted relatively limited interest: University Training Schools. This model was originally set out in the 2010 Schools White Paper but, as far as I can establish, only the University of Birmingham and the Institute of Education have so far taken this path. The latter project seems rather under wraps and this is the only explicit link I can find on the IoE’s own website, though it is also mentioned in this TES article. (I found a reference in Paragraph 43 of Oxford University’s Access Agreement for 2012/13 to ‘development of a University Training School as a laboratory school, once the procedures for developing these schools are clarified’ but this seems to have fallen out of the latest 2013/14 Agreement.)

It is quite likely that many potentially interested and eligible universities have already backed a different model and are reluctant to expand their portfolio at this stage.

Some – such as Warwick University – will be relying on other initiatives to secure a stronger share of the best undergraduates. In Warwick’s case that role is fulfilled by IGGY.

This comparatively limited interest is despite the fact that a capital budget of approximately £6m is available for each project, plus annual recurrent costs of around £4,000 per student in Exeter and £5,115 in Lambeth, London (according to the DfE’s ‘ready reckoners’) not to mention the unspecified sums available in development and outreach grants, or any other supplements made available.

It is not clear how much of a university’s own money would be needed for such a project but one might expect that any cost would be attributable mainly to the staff resource needed to develop and launch the project and then provide steady-state input, including the specified contribution to the teaching and support of students.

That would be a tidy sum no doubt, but surely covered substantively by a development grant and the recurrent funding available. (Set in this context, the apparent decision to withdraw a development grant from new applicants seems rather puzzling.)

.

From Kew Gardens courtesy of Gifted Phoenix

From Kew Gardens courtesy of Gifted Phoenix

.

Assessment of the model

 

Selection

16-19 maths free schools are based on the twin pillars of selection and specialisation.

The arguments for and against selection are well-rehearsed and I will not repeat them here. It seems that selection at age 16 is somewhat less contentious than selection at age 11 (with selection at 14 a largely untested assumption).

Nevertheless, most of the arguments against (and for) selection remain in play regardless of the age at when that selection takes place. We can see this writ large in current debate about fair access to university and its impact on social mobility.

It seems unlikely, therefore, that selective universities would harbour an ideological opposition to selection at age 16.

The pitch of the selection is critical. The description of the Government’s policy intention would suggest a cadre of highly selective institutions, though of course that depends ultimately on the number of candidates who apply and, of those, what proportion can satisfy the admissions criteria.

Some aspects of those requirements are currently unclear, even for the school at the most advanced stage of development. For example, we know nothing of the planned aptitude test at KCLMS.

It is clear that their GCSE requirements are not as exacting as they might be, in that they do not require A* grades in maths and physics or a compulsory pass in English.

The latest 2011/12 statistics suggest that 20.2% of students achieve an A*/A grade in mathematics while almost 47% manage this in physics. Given the similarity between the subjects, it is fairly likely that the proportion achieving this level in both subjects (or in maths and combined science) is also likely to be fairly close to 20%.

This places the pitch of selection on a par with the traditional assumption for grammar schools (though the reality is now far different and highly differentiated).

There is an obvious trade-off here between excellence and equity. If selection is pitched too highly, it will become impossible to recruit sufficient students from disadvantaged backgrounds, because high attainment is found disproportionately amongst those from comparatively advantaged backgrounds. As I have suggested, this could mean that the provision is unfairly monopolised by the middle classes.

On the other hand, if it is pitched too low, students will be admitted who are not the very highest achievers and so are relatively less likely to achieve the A level grades they need to secure places in the most competitive university maths departments.

Gifted educators know that this issue boils down to the critical distinction between attainment and ability.

These schools need to find the right blend of admissions arrangements such that they can recruit:

  • A critical mass of the highest achievers from a variety of backgrounds, ideally giving preference to those whose current institutions do not offer high quality post-16 maths education, rather than the products of selective and independent schools; and
  • An even more critical mass of students with demonstrated mathematical ability which may not yet have been translated into high achievement, especially those whose underachievement is attributable – at least in part – to a relatively disadvantaged background.

KCLMS’s aptitude test will be critical in achieving this outcome, as will their decision whether or not to give priority admission to recipients of the Pupil Premium. It will be important that they and Exeter subject their draft admission criteria to proper ‘stress testing’ before they are adopted.

.

Specialisation and Acceleration

The debate over specialisation is less polarised. Historically there has been argument that the typical A level student experiences a rather narrow curriculum compared with his peers in many other countries, including several of those perceived to have the most successful education systems.

The trade-off between breadth and depth is discussed in Ofqual’s Report on International Comparability.

But the Consultation on A level reform did not enter this territory other than in relation to AS levels, arguing that the majority view is that A levels are broadly ‘fit for purpose’.

The specific issue in this context is that students attending these schools are likely to have an even narrower curricular experience than their peers in other English schools and colleges.

If the KCLMS precedent is followed, they will have an extremely constrained choice of A levels – indeed no choice at all – compared with what would be available in a typical sixth form, even in a small rural school.

There are references to curricular provision beyond maths and physics in the KCLMS plans, but it is not clear how they will be implemented in practice, beyond the option of an AS Extended Project.

It has to be open to question whether a small sixth form containing 60 students in each year group, all taking the same three A level choices, is the optimal solution for many students who, as a consequence, will not be exposed to ideas and perspectives from peers experiencing an entirely different subject context.

There will be limited opportunity to bring out the inter-disciplinary connections that are so often of interest to gifted learners, to undertake cross-curricular collaborative learning with peers who can bring to bear strength in other subject areas.

This seems an artificial constriction which may make the KCLMS option unattractive to some students, especially those who are ‘all-rounders’ with strength in maths and other subject areas. It is not necessarily a given that these students will be weaker mathematicians than peers with just that one string to their bows.

Moreover, the KCLMS proposal is guilty of a different kind of narrowness in that it is avowedly anti-acceleration, so ignoring opportunities to utilise the close relationship with a university to enable school-age students to pursue undergraduate study.

This reflects a strong strand of thinking in parts of the UK maths education community which believes that acceleration is most definitely not in the best interests of students.

It is not the position I would take, which is that acceleration (faster pace) done properly can be combined effectively with enrichment (greater breadth) and extension (more depth; more problem-solving), and that the proportions of each should reflect different students’ needs. (There is not space here to unpack what ‘done properly’ means, but most gifted educators will be familiar with the arguments.)

The KCLMS approach will probably be unattractive to some of the very highest achieving young mathematicians, who will see this as placing an artificial cap on their progress. It will also mean that KCLMS is very different indeed to some comparable institutions in other parts of the world where accelerated study is actively encouraged.

(I note in passing that it is as yet unclear whether these schools will admit already-accelerated students aged under 16.)

.

A Network of New Schools or a Broader National Network?

One might reasonably question whether setting up a tranche of a dozen or so schools at a capital cost of £72m and an annual recurrent cost in steady state of approximately £6.5m (assuming 12 schools with 120 students each and an average annual per student recurrent cost of £4,500) is the most efficient strategy for increasing the supply of high achievers in maths.

Especially since the benefit under this model is largely confined to an annual cohort of around 720 students (12 x 60) assuming there are 12 schools all the same size as the first two.

In order to roll out the same model, further funding tranches of this magnitude would be required for every additional 12 schools added to the network – there would be few if any economies of scale.

It is likely that this model was adopted because: the Government wanted to increase the stock of free schools; the available capital funding could not be diverted to cover running costs; and it was felt that the infrastructural work involved in building the new schools would itself have a positive impact on economic growth.

Also, perhaps, because, it is ideologically committed to a ‘bottom-up’ distributed model rather than a ‘top down’ prescriptive model – and is reluctant to entertain the possibility that there might be an optimal ‘middle way’.

It would be quite wrong to criticise the current programme at this early stage because we have no evidence of its impact, other than on the grounds that the number of beneficiaries will be comparatively small.

It may eventually be demonstrated that the positive impact on students is so marked that the programme is good value for money despite the heavy outlay.

But, if we were given a development budget equivalent to the cost of one school (£6m), an identical annual running cost budget of £6.5m per year and a blank sheet of paper, what design principles might we establish to underpin a more efficient and fully scalable approach?

One might begin with the core purpose of creating and sustaining a national network designed to support all students in state-maintained schools and colleges with the potential capacity to achieve, say, at least grades AAB in three of the target A level subjects plus a STEP paper grade of 1 (very good) or S (outstanding).

Such support would be available from Year 9 at the latest and ideally from Year 7. From Years 7 to 9 it would be light touch and provided to a relatively broad cohort, in recognition of the difficulty of predicting future performance at such an early stage.

But, from Year 10, it would be concentrated on a smaller group of future high achievers. This would include existing high attainers, but would also give priority and additional intensive support to learners whose potential is significant, but is unfulfilled as a consequence of socio-economic disadvantage.

This national network would need to draw on the co-ordinated strength of the many national bodies already active in this field, including the likes of Nrich, the NCETM and MEI’s Further Mathematics Support. They would be drawn into a powerful coalition, prepared to sink their differences in pursuit of this common cause. (Those receiving Government funding might have it made conditional on their constructive involvement.)

The network would aim to reach every state-maintained secondary school and post-16 institution, and to draw directly on the expertise within the widest range of institutions which have it to offer, including specialist academies, outstanding schools with an old-style maths specialism, national teaching schools, independent schools and post-16 institutions.

It would be developed on ‘flexible framework’ principles, combining a set of challenging common core expectations and light touch accountability with sufficient autonomy for participating institutions to innovate and to meet the very different needs of their students.

The services provided and co-ordinated through the network might include:

  • Outreach by the strongest university, college and school maths departments in each region, regardless of the categorisation of those institutions.
  • Extensive online learning provision, for use in class and via independent learning, again drawing on the combined expertise of all national, regional and local partners. This would be free at the point of delivery and would be designed on social network principles, encouraging students to learn with and from each other.
  • Support from an undergraduate of postgraduate mentor, provided face-to-face in the case of those from disadvantaged backgrounds.
  • Additional support to raise the aspirations of students from disadvantaged backgrounds and to equip them with the social and cultural capital necessary to compete for places at the most competitive universities.
  • High quality professional development and support for host schools and colleges and lead mathematics and physics teachers within them

In addition, a small core of schools and colleges – some academies and free schools, some not, some independent – might be identified as post-16 centres of excellence and funded to admit the most promising students from disadvantaged backgrounds.

In the short term there would be ‘quick win’ interventions in the form of direct support for disadvantaged learners across Years 12 and 13.

The Government would ensure that all appropriate policy connections were made – whether with wider support for maths education, academically able pupils, fair access to higher education and so on – to ensure that all are mutually supportive and that benefit from the whole is greater than the sum of its parts.

And of course the whole caboodle would be rigorously evaluated, both formatively and summatively. Success would be judged against achievement of a few rigorous performance measures.

.

The Bigger Picture

As we approach the 2013 Budget, there are many signs that we have emerging consensus on the importance of investment in human capital. Witness, for example:

  • The Heseltine Report on growth ‘No Stone Unturned’, the bulk of which has been accepted by the Government

No doubt there are many more.

But, with the honourable exception of the CBI (which is not as explicit as it might be on the point) none of these recognise the substantial benefits that would accrue from more targeted investment in our school-age high achievers.

To give the Government credit, the 16-19 maths free schools programme shows that they are alive to these arguments, even if only in a relatively narrow STEM-related context.

But it is worth pausing to consider whether a network eventually built around a small set of selective post-16 institutions is the optimal approach.

Assuming that new free schools are a ‘non-negotiable’ it might be preferable to start with the network and drop the schools into it, rather than starting with the schools and waiting for them to build the network from the bottom up.

There are lessons to be learned from the careful study of similar provision in jurisdictions like Hong Kong, Singapore, South Korea, Taiwan and Israel, all of them featured in earlier posts on this blog. In these jurisdictions, the ‘elite’ schools are typically nodal points in a much wider mesh of provision rather than ‘stand-alone’ providers with outreach capacity.

An evaluation of the maths 16-19 free schools pathfinder project might usefully incorporate that comparative dimension, while also reflecting the current predilection for randomised control trials.

Given the recent designation of the Education Endowment Foundation as a more generic ‘what works centre’ for education, it may now be for that body to commission the appropriate study.

.

GP

March 2013