Primary assessment in England - The National Middle Schools` Forum

1
NMSF Paper – May 2017
Consultation - Primary assessment in England
Considering the NMSF response
Executive Summary
The National Middle Schools’ Forum welcomes the consultation on the future of Primary Assessment
and hopes that there will be an opportunity to develop a system of assessment and accountability
that properly recognises the unique contribution of three systems within our increasingly diverse
education system. The middle school sector educates a significant number of children - there are
currently over 55,000 pupils on the roll at our middle schools. The national system of assessment
and accountability needs to reflect properly the nature of our systems. The recent difficulties
experienced in applying the coasting schools definitions to middle schools have served to highlight
the need for an agreed framework for middle school accountability. In the words of one of heads,
“we need to stop trying to force a square peg into a round – and redesign the hole with the nature
of the peg in mind.”
We agree with the points raised in the section ‘School types and assessment’ on page 21 of the
consultation paper:
An accountability system for middle schools should:
a) Cover the time when a pupil is at a middle school - that is from the point they enter the
school until the time they leave.
b) Avoid measures which span two types of school – where accountability is shared across a
key stage. Middle schools cannot be accountable for time spent at another school.
Accepting a system for middle schools such as that proposed in the consultation paper that
measures progress from the KS1 tests fails these two tests when applied to middle school systems.
Middle schools cannot be accountable for progress across KS2 when over half of the key stage is
spent in first schools.
We would propose, rather, that we work together with DfE and partners to develop a system which
builds on the Forum’s own work on middle school accountability and meets the two tests above by
providing:
i) A baseline assessment when pupils enter their middle school
ii) An assessment of progress to the end of year 8
Such an arrangement would build on the structure of the revised national curriculum that divides
the KS2 programmes of study into upper and lower KS2. Only then can middle schools properly
account for the progress pupils make while through their time in the middle schools.
2
Background – Consultation document and middle schools
The consultation paper is available at bottom of this web page https://consult.education.gov.uk/assessment-policy-and-development/primary-assessment/
See Page 21
Q10. If we were to introduce a reception baseline to enable the creation of reception to
key stage 2 progress measures for all-through primaries, what would be the most effective
accountability arrangements for infant, middle and junior schools’ progress measures?
School types and assessment
The introduction of a new assessment in reception as a baseline for measuring progress
would have an impact on infant, junior and middle schools. We want to ensure that we
continue to have the most appropriate accountability arrangements for these schools. In
2013, we committed to developing a national progress measure for infant schools (from
reception to key stage 1) and retaining progress measures for junior and middle schools
(from key stage 1 to key stage 2). Subject to the questions set out in the preceding sections,
we will need to reconsider the best accountability arrangements for these types of school.
In keeping with the policy intention of progress measures covering the time within which
pupils are in a school, the most logical measures for infant schools would be reception to key
stage 1 and, for middle and junior schools, would be to continue with key stage 1 to key
stage 2. This would mean that these schools would be judged on a different basis from allthrough primary schools and so would need to be compared against each other, rather than
all other schools with key stage 2 provision. For example, pupils in junior schools would not
be compared against all pupils nationally with similar starting points, but rather against only
those pupils who were in infant schools at the end of key stage 1. This could make it more
difficult for parents to compare results.
The alternative would be to hold infant and junior schools to account using a single
reception to key stage 2 progress measure, encouraging greater collaboration between
infant and junior schools. This, however, presents its own difficulties, as it would involve
holding both schools to account for the progress made across the 7 years, rather than just
the time the child spends in their school.
We are keen to explore these issues and would welcome views on the approach for infant,
junior and middle schools.
A) Introduction - The points raised in the consultation paper:
“In keeping with the policy intention of progress measures covering the time within which
pupils are in a school...” (Page 21)
For middle schools this stated policy intention would imply a system that measures progress from
Year 5 to Year 8 in most cases.
3
“….the most logical measures for infant schools would be reception to key stage 1 and, for
middle and junior schools, would be to continue with key stage 1 to key stage 2.”(page 21)
Retaining KS1 and using it to measure progress from KS1 to KS2 for middle schools makes no sense
as it is incompatible with the stated policy intention. Retaining KS1 would retain the current
difficulties where responsibility for KS2 is split between two different types of school –the very
problem used as a reason not to support baseline to KS2 in infant / junior systems in the next
paragraph of the paper:
“The alternative would be to hold infant and junior schools to account using a single
reception to key stage 2 progress measure, encouraging greater collaboration between
infant and junior schools. This, however, presents its own difficulties, as it would involve
holding both schools to account for the progress made across the 7 years, rather than just
the time the child spends in their school.”(page 22)
Taking these points from the consultation paper together then and applying this logic to middle
school systems - an accountability system for middle schools should:
a) cover the time when a pupil is an a middle school
b) avoid measures which span two forms of school – where accountability is shared
National Middle Schools’ Forum agrees with these two statements, which we take to be the
starting point for any system of assessment for middle schools. This current review of assessment
arrangements represents an opportunity to put in place a system that respects the unique
contribution of three systems to our increasingly diverse education system.
B) Background - The work of NMSF Developing a robust accountability framework for
middle schools
The Forum has for some time worked with Ofsted and other agencies to develop a framework for
accountability for middle schools that has enabled our schools to present robust and reliable
measures to parents, inspectors and other interested parties.
2011 – 2015 Benchmarking. For a number of years the Forum provided benchmarking data against
which schools could establish the profile of their intake and demonstrate pupil progress to the end
of Year 8. Middle School pupils were tested at intake and at the end of Year 8 using the optional SAT
tests, frequently using external marking or moderation and submitted their results. The analysis of
these results enabled schools to benchmark their results against a large sample. Comparative data
over the five year period of our benchmarking work suggests that this was one factor in the
improving performance of middle schools, both in core and foundation subjects.
See Annexe 1 – summary of benchmarking work 2011 to 2015
2015 onwards – Partnership with GL Assessment. Following the demise of national curriculum levels
we sought to develop of partnership with a provider of nationally benchmarked tests that would
cover the middle school age range and enable us to benchmark first and middle school achievement
against a large and robust national sample. We have been developing a partnership with GL
Assessment to establish a middle school accountability framework using their Progress Tests across
KS2 and KS3. This enables individual middle schools to:
4


Establish ability on intake through tests in Year 4 or Year 5
Monitor progress through the middle school years through the point of transfer to High
School.
Evidence from recent Ofsted inspections and the first round of Coasting Schools with RSCs confirm
that this framework provides a robust reliable basis of accountability that has general acceptance.
c) Principles that should underpin a system of accountability suited to three tier systems
1) Middle schools expect to be accountable for the progress of their pupils during
their time in the school – that is progress from point of entry to point of exit.
See Statutory Guidance – School Causing Concern
Arrangements for middle and other schools
The majority of middle schools are deemed to be secondary schools but due to the
age range of pupils are subject to the coasting definition relating to key stage 2,
rather than the coasting definition relating to key stage 4. A pupil may, however,
only have attended a middle school for a short time before they took the key stage 2
tests and may still have a number of years left at the school. For this reason, RSCs
will give consideration to the wider context when a middle school falls within the
coasting definition. This will include giving consideration to the progress made by
pupils from the point of entry to the middle school to when they leave, which may
be demonstrated by robust, and where possible externally benchmarked, school
data. (Page 15)
Available here - https://www.gov.uk/government/publications/schools-causing-concern--2
2) There is no valid and reliable way in which middle schools can be accountable
for progress during KS2 when over half of KS2 is spent in first schools.
To illustrate this point we could compare two middle schools with identical KS2 results in the
chart below, both just below the floor target. In the first case the KS2 result could indicate
an ineffective middle school, but equally well this could indicate the second case where an
ineffective first school combined with an effective middle school. Both middle schools
achieve identical outcomes at KS2 – so no valid or reliable judgement about their
effectiveness can be drawn from their KS2 results.
5
3) KS1 results within three tier systems are an unreliable starting point for
measuring progress – the KS1 results in first schools show a markedly different
profile of results from that in all through primary schools.
See Annexe 2 below - which summarises research on KS1 measures in three tier systems.
Data on progress from KS1 to KS2 within three tier systems is doubly unreliable:
i) Middle schools are only responsible for part of the progress a child makes during
KS2
ii) Pupils in first schools are given higher KS1 scores than comparable pupils in
primary schools. This has the effect of making progress from KS1 to KS2 appear
lower in three tier systems.
6
D) Proposals for a middle school assessment and accountability system
When the revised national curriculum was under discussion we hoped that the division of the KS2
programmes of study into upper KS2 and lower KS2 would lead to the development of a system to
check progress part way through KS2 – perhaps through the development of optional tests. This
would have the potential to address long-standing national concerns about progress in lower KS2
and enable the development of an appropriate system of accountability for three tier systems. This
seemed to be an opportunity missed when this idea was rejected.
There would seem, then, to be two possible avenues to the development of a system of assessment
and accountability for middle schools:
1) The development of materials or tests appropriate for the end of Year 4 or start of Year 5
that would also be available to all through primary and junior schools to use on an optional
basis.
2) The commissioning of a test to be taken at the end of Year 4 or the start of Year 5 in
middle school systems only from an appropriate commercial provider of progress tests.
One of these options, in combination with the commissioning of an appropriate test to be taken at
the end of Year 8, would provide an appropriate framework for middle school accountability. These
options would also provide, for the first time, a measure for the accountability of first schools, for
whom the use of KS1 outcomes alone is equally problematic.
Many middle schools took part in the ‘Two Year KS3 Project’ – and continue to cover the KS3
curriculum in Year 7 and 8. This is possible because in middle school systems pupils can begin their
KS3 programme of study in the second half of the Summer term in Year 6, and avoid the deficits in
progress associated with transfer at age 11. With this in mind, it would similarly be possible to
develop test materials for use at the end of Year 8, covering the KS3 programme of study, that could
be useful to secondary schools in Year 9 if they were available on an optional basis.
As you would expect NMSF benchmarking data reveals that pupils in middle schools make good
progress in KS3. It is frustrating that in current arrangements, that focus inappropriately on KS2
outcome alone, this good progress in Years 7 and 8 goes unacknowledged. Under the system we
propose there could at last be a proper appraisal of the full progress pupils make during their time in
their middle school.
7
Annexe 1 – Background – Summary of NMSF Benchmarking
KS3 Benchmarking – Looking at trends over the last six years
Between 2011 and 2015 the National Middle Schools Forum provided an
analysis of Year 8 test data to enable schools to benchmark their own data
against a large national sample.
This paper considers trends shown in Year 8 attainment in English, Maths and Science over this
period. For Maths and Science the results reflect the achievements of over 7,000 pupils each year
(and from over 60 contributing schools each time). Not all schools report a single English level and so
the sample size here is slightly smaller – in 2014 the results from 47 schools are included (reflecting
the achievement of over 5,000 pupils).
English
% achieving
each level
National
Year 9 2007
2011
2012
2013
2014
2015
Below
Level 4
Level 4
Level 5
Level 6
Level 7
Level 8
APS
13
13
41
25
8
0
33.5
4.9
N/A
0.5
1.4
1.4
13.6
40.9
35.6
4.9
0.1
34.3
2.1
10.5
8.7
12.4
38.4
37.3
40.1
41.0
40.1
37.7
9.0
11.3
7.2
0.1
0.1
35.0
35.9
36.0
8
(We did not collect English levels in 2012 – choosing rather to collect separate reading and writing levels. In
2013 it was decided to collect both single English levels and reading and writing, because it became clear that
practice differed in individual schools.)
Maths
% achieving
each level
National
Year 9 2007
2011
2012
2013
2014
2015
Below
Level 4
Level 4
Level 5
Level 6
Level 7
Level 8
APS
10
14
20
27
21
8
36.7
5.7
2.7
2.7
2.7
1.9
11.8
10.8
9.5
9.8
8.9
24.3
23.8
21.6
20.7
20.4
33.5
31.4
32.9
32.2
30.4
20.7
25.0
26.0
27.0
28.3
4.1
5.9
7.1
7.6
9.8
36.8
37.9
38.5
38.6
39.2
Below
Level 4
Level 4
Level 5
Level 6
Level 7
Level 8
APS
10
17
32
26
15
0
34.3
2.7
0.5
0.6
0.7
0.3
7.9
7.7
6.8
6.9
6.9
35.0
36.0
32.8
32.9
29.5
43.4
43.4
48.3
45.4
46.9
11.0
12.2
11.4
14.1
16.0
0.0
0.0
0.1
0.0
0.2
36.1
36.5
36.8
36.9
37.3
Science
% achieving
each level
National
Year 9 2007
2011
2012
2013
2014
2915
9
Conclusion
Progress at KS3 continues to show an improving trend in all three subjects over this period.
The only remaining point of national comparison remains the data from the last National Curriculum
SATs tests for Year 9 pupils in 2007. These charts confirm what we have known for some time:
a) At the end of Year 8 middle school pupils achieve results which surpass those of Year 9
pupils in 2007.
b) Middle school pupils make better progress in KS3 than pupils in other forms of schooling.
10
Annexe 2 – The unreliability of KS1 to KS2 progress data in middle school systems.
a) We worry about teachers inflating results; we should worry more about depression of
baseline assessments.
In this research published by Mike Treadaway of Datalab he poses the question ‘Why is Key
Stage One assessment so different in infant and primary school systems?’
As this graph shows pupils with similar Foundation Stage Profile results appear to make
lower progress in infant/junior systems (and first/middle school systems) across the ability
range:
This is explained, he concludes, by the divergence in Key Stage One scores in the two forms
of schooling after the introduction of teacher assessment in 2003:
Before 2003, infant schools achieved only slightly higher Key Stage One scores than primary
schools, but after teacher assessment was introduced, their scores started to diverge
strongly. Pupils in primary schools are consistently given lower Key Stage One scores, than
those in infant and first schools.
b) NMSF – evidence from intake benchmarking
11
This same difference between KS1 baseline scores is evident in National Middle Schools
Forum benchmarking data for pupils entering middle school:
September 2013 Intake
Number of pupils in First School sample
KS1 Data from First Schools
National KS1 2011
5375
Reading
Writing
16.8
15.6
15.8
14.4
Maths
16.5
15.7
KS1 - Comparing First School and National APS
Maths
National KS1 2011
Writing
KS1 Data from First Schools
Reading
13.0
14.0
15.0
16.0
17.0
18.0
September 2014 Intake
Number of pupils in First School sample
KS1 Data from First Schools
National KS1 2012
5719
Reading
Writing
16.8
15.7
16.0
14.7
Maths
16.6
15.9
KS1 - Comparing First School and National
APS
Maths
National KS1 2011
Writing
KS1 Data from First Schools
Reading
13.0
14.0
15.0
16.0
17.0
12
The lowered profile of KS1 scores in primary schools compared to those for first schools means that
data concerning progress from KS1 to KS2 within middle school systems are not directly comparable
national data which is largely based on data from primary schools.
The evidence clearly suggests that pupils of comparable ability in first schools are given higher KS1
scores than those in primary schools – this distorts data on pupil progress within middle school
systems, suggesting lower progress than is actually the case when compared to national figures
which are largely based on progress in primary schools.