Supplementary material submitted to Education
and Skills Select Committee
September 16th 2003
Professor John Coldron
Sheffield Hallam University
Contents
1. Findings concerning the choice
process - schools and parents
1.1 Use
of performance tables and likelihood of opting away from local
school:
1.2 Do
parents demote their favourite school?
1.3 Likelihood
of parents being offered their favourite school:
1.4 On
the characteristics of schools that parents class as their favourite
school.
1.5 Characteristics
of parents who appeal:
2: LEA procedures for expressing
a preference
2.1 How
many LEAs use equal weighting of parents' expressed preferences?
3. The
management of preferences in wholly selective areas
3.1 Selective
areas
3.2 Problems
of admission management in selective areas:
3.3 Ways
in which selective LEAs try to solve the problem:
3.4 Adjudicator
rulings:
4. Management of the problem in areas
that are not wholly selective but have wholly selective schools
within their area
4.1 LEA
practice:
5. Access to tests for places at
a selective school
5.1 Opting
in or opting out?
5.2 Possible
problems with non-universal systems:
6. Evidence as to the educational
benefits of selection
6.1 Evidence
from Schagen and Schagen and PISA
7. Proposals and recommendations
7.1 The
need to accommodate different solutions
7.2 Issues
to be considered
7.3 Proposals
and recommendations
1. Findings concerning the choice process - schools
and parents
1.1 Use of
performance tables and likelihood of opting away from local school:
There are a number of relevant findings
in the nationally representative survey conducted by the Office
for national Statistics for our joint DfES study (Flatley and
Williams 2001) into the experience of parents. The figures are
for 1999/2000.
The findings provided further evidence that parents
consider a whole range of criteria of choice and that the academic
performance of the school is only one. For example a minority
of parents (39%) used performance tables to find out about schools.
However parents among whom the mother had a degree or above were
nearly twice as likely to consult them as those among whom the
mother had no educational qualifications.
We also found that parents in London, those with
higher educational qualifications and those in occupational classes
I and II were more likely to opt away from their nearest school
(see Table 1). This shows that parents who live in London were
six times more likely to apply for a place outside their own LEA
area than parents who live in a Shire authority and that parents
where the mother had never had paid employment were three times
less likely than those in Social Class I or II to have applied
outside their own LEA; and where the parent had no educational
qualifications they were half as likely (two times less likely)
to have applied outside their own LEA (see Table 1).
Table 1 Likelihood of parents applying for a place
in a school outside their own LEA area (based on odds ratios from
logistic regression)
| Characteristic
| Odds ratios
| 95% confidence intervals
|
| | Lower
| Upper
|
| Highest educational qualification
| | | |
| Degree or equivalent or higher
| 1.0
| .. |
|
| Other qualifications
| 0.5
| 0.4 | 0.9
|
| No qualifications
| 0.4
| 0.2 | 0.7
|
| | |
|
| Social class of mother
| | | |
| I & II
| 1.0
| .. | ..
|
| III non-manual
| 1.0
| 0.7 | 1.5
|
| III, IV & V manual
| 0.9
| 0.6 | 1.4
|
| Never worked
| 0.3
| 0.2 | 0.6
|
| | |
|
| Parental LEA type
| | | |
| London borough
| 6.3
| 4.2 | 9.4
|
| Metropolitan authority
| 1.3
| 0.8 | 2.0
|
| Unitary authority
| 2.6
| 1.8 | 3.8
|
| Shire authority
| 1.0
| .. | ..
|
| Nagelerke R2
| 0.1
| | |
1.2 Do parents demote
their favourite school? Members were interested
in whether there was a difference between the school that parents
stated on the application form as their first preference and the
school they would 'really' have preferred. The ONS analysis covered
this aspect of parental choice to some extent. Parents were asked
if there were any state schools they had not applied to but would
have preferred their child to attend. Eight per cent of parents
reported that there were. In addition the survey collected the
names of all schools to which parents applied for a place. Parents
were asked which of these schools they most wanted their child
to attend. For the purposes of analysis this school was referred
to in the report as the parent's favourite school. It is
therefore possible to infer what proportion of parents demoted
their favourite school to a lower preference in the ranking
of their application form. The survey found that while 92% gained
entry to the school they had put as their first preference 85%
reported that they had gained their favourite school. This
indicates that about 7% had not put their favourite school as
their first preference. The results are presented in Table 2.
Table 2: Outcome measures by cohort
| % offered a place in
| Entry cohort of selected child
|
| Sept 1999
%
| Sept 2000
%
| Total
%
|
| In a preferred school
| | | |
| Favourite school
| 89 | 81
| 85 |
| First preference school
| 93 | 91
| 92 |
| Any school for which preference expressed
| 97 | 95
| 96 |
| | |
|
| In a school for which no preference stated
| 3 | 5
| 4 |
Weighted base
| 1192
| 977
| 2170
|
1.3 Likelihood of parents
being offered their favourite school:
It was found that there was not a strong relationship between
the background characteristics of parents and the likelihood of
being offered a place in their favourite school. However
parents in London were an exception and were the least likely
to be offered a place in their favourite school (an odds ratio
of 0.3:1.0 compared with the reference category of Shire authorities).
While nationally 85% of parents were offered a place in their
favourite school in London only 68% receive such an offer. Nationally,
4% of parents were offered a place in a school for which no preference
had been stated. Nevertheless, among these parents around half
reported they were satisfied with the school they had been offered.
1.4 On the characteristics
of schools that parents class as their favourite school.
Six in ten of the favourite
schools (as defined earlier) had higher than average GCSE
performance scores than their LEA average. When compared to national
figures for the proportions of pupils eligible for free school
meals, 46% of favourite schools fell within the two lowest
national quintile groups (i.e. in the 40% of schools with the
lowest proportions of students receiving free meals). Fifteen
per cent of the favourite schools were in the 20% of schools
with the highest proportions of students receiving free school
meals.
When the odds of all the different factors in
combination are examined, parents who had no previous experience
of choosing a secondary school, were owner occupiers, lived in
London boroughs, had a degree level qualification or above and
were in Social Class I and II had the highest odds of choosing
a favourite school with a high GCSE performance score.
They were three times more likely to do so than those parents
in the reference category (who differed from them in having previous
experience of choosing a secondary school and in living in a Shire
authority). Parents who had the lowest odds of choosing a secondary
school with a high GCSE performance score had previous experience
of choosing a secondary school, lived in social sector rented
accommodation, lived in a Shire authority, had no qualifications
and were parents among whom the parent had never worked. This
group of parents were about eight times less likely than the reference
group of parents to choose a favourite school with a GCSE
performance score above the LEA average.
1.5 Characteristics
of parents who appeal: Five percent of
parents in the nationally representative survey reported that
they had appealed (n=129). In general, these parents did
not vary by background characteristics. In other words there was
no greater likelihood of a middle class parent appealing than
a working class parent. In our work for the DfES project on Appeals
panels (Coldron et al 2002) panel members reported their impression
that there were now greater numbers of working class appellants.
We also found that there was no greater likelihood of a middle
class parent winning an appeal i.e. the Social Class of an appellant
bore no significant relation to whether the appeal was successful
or not (n=317).
2: LEA procedures for expressing a preference
2.1 How many LEAs
use equal weighting of parents' expressed preferences? The
question was raised as to how many LEAs used an equal weighting
procedure. The nationally representative survey (Flatley and Williams
2001) found that 66% of parents were asked by their LEA to list
their preference for school, on their application form, in rank
order.
In Stage One of the same project (Williams et al
2001) we analysed all eligible LEAs in England (n=141) and categorised
their procedures for allowing parents to express a preference.
This data relates to the procedure for September 2000 entry. Finding
a means of categorising LEA modes of practice which captures the
complexity that exists, whilst ensuring that a useful and usable
typology was developed presented several challenges. In some areas
LEA admission procedures applied to such a small proportion of
schools that describing the LEA mode of practice did not provide
a meaningful description of the procedures that many parents experienced
there. The procedures found in the composite prospectuses were
categorised into four types.
Type 1: Multiple preferences - Rank order
The parent is invited to name a number of schools
as preferences in rank order and the LEA admission authority attempts
to allocate them a place at their first choice school. Though
the first choice takes precedence in the majority of cases it
is not guaranteed that a parent's first choice will take precedence
over someone else's second choice. There are circumstances where
a second choice takes precedence, particularly if this means a
pupil would be spared a long or difficult journey to an alternative
school. Sometimes a multi-stage process is involved where parents
may be asked to enter the next stage of the admission procedure
and identify further preferences if they have been unsuccessful
with their initial preferences.
Type 2: Invitation to accept a designated school
or choose another
Parents are notified of a place allocated at a school
chosen by the LEA on the basis of the general admission arrangements
of the authority and invited to confirm that particular school
as their expressed preference or to name one or more alternative
preferred schools. The parent must confirm in writing that the
allocated school is acceptable otherwise it is treated as a non-preference
and no place is reserved. This procedure is most commonly used
in those areas where a catchment or priority area system is in
place. How the consequences of this are handled by LEAs varies
considerably. Some LEAs state explicitly that not putting the
designated school down as first preference may mean that a place
is not available at this 'local' school should they be unsuccessful
in gaining a place at a different first preference school. Other
LEAs state explicitly that parents who put their 'catchment 'school
down as a second preference who do not get their first preference
will not be treated any less favourably than a parent who chose
the catchment school as their first preference.
Type 3: Multiple preference equal weighting
Parents are asked to express multiple preferences
and each preference is given equal weighting. LEAs then allocate
to one of those preferred schools on the basis of their general
admission arrangements. This is often in context where the admission
system is complex involving a number of admission authorities.
In such circumstances the procedures can seem to be complex unless
there is a common admission timetable and a common application
form as is now required by the new Code (DfES 2003).
Type 4: Single preference system
Parents are invited to express a single preference
and if this proves unsuccessful they may apply to alternative
school(s) in a second stage of dealing with preferences.
Table 3: Frequency of procedures for expressing
a preference (Sept 2000 entry)
| Type of system
| No. of LEAs
| % of LEAs
|
| Multiple preference rank order
| 105
| 74
|
| Invitation to accept designated school or express a preference for other school(s)
| 18
| 13
|
| Multiple preference equal weighting
| 7
| 5
|
| Single preference
| 11
| 8
|
| Total
| 141
| 100.0
|
3. The management of preferences in wholly selective
areas
3.1 Selective
areas: Although only a minority, selective
admissions present very different issues of management for both
school admission officers and parents. The official definition
of a selective LEA is one where over 25% of the pupils attend
selective grammar schools. There are other areas such as Chelmsford
and Colchester in Essex, or Ripon and Skipton in North Yorkshire
where all of the schools reasonably available to parents are organised
as either selective grammars or secondary moderns. But because
these are sub-areas within larger LEA districts the LEAs are not
classified as wholly selective. All wholly selective areas use
some version of a standardised test of general ability administered
at 11+. In practice this was either one produced by an independent
research organisation such as the NFER or one developed within
the LEA itself. We deal in turn with two aspects of admissions
in these areas -the first being the allowance of two first choices
and the second being access by pupils to selective tests.
3.2 Problems of
admission management in selective areas: The
advent of open enrolment highlighted a tension between the new
inclusive principle that gave parents the theoretical right of
attendance at any school of their choice and the older exclusive
principle of selection. This, and the requirement to take account
of parents' expressed preference creates a difficult issue for
parents who wish to apply for a selective place if, at the time
of stating their preference, they do not know if their child is
eligible for a selective school place. If they express a first
preference for a place at a selective school and their child does
not reach the required standard (or too many do) it is possible
that they would not get their preferred non-selective school because
that school may already have reached its admission limit through
the allocation of first preferences.
3.3 Ways in which
selective LEAs try to solve the problem: In
the event of a child not gaining admission to a first choice selective
school some wholly selective areas manage this difficulty for
parents by allowing the next non-selective school preference to
be deemed a first choice equal with those other parents who had
expressed this as their actual first choice. Thus, parents who
apply to selective schools are advantaged over other parents who
for whatever reason do not apply. We found two variations of this
default process. One is to maintain separate selective and non-selective
school preference lists deciding which to use when the child's
11-plus result is known. The second way is to require parents
to express preferences for selective and non-selective schools
on a common form. If it turns out that the child is deemed ineligible
for a place at a selective secondary school the highest preference
non-selective school becomes, by default, the first preference.
These systems have either been accepted by parents and schools,
or at least have operated without any effective objection, over
a number of years.
3.4 Adjudicator
rulings: While the 1998 Education Act
reaffirmed both open enrolment and then existing systems of selection
the legislation also created the Office of the Schools Adjudicator
thereby creating a process for challenging the admissions practices
of schools as admission authorities and of LEAs. The Schools Adjudicator
has ruled against the default procedures described above in two
important determinations one concerning Wirral and the other Torbay.
In Wirral the adjudicator decided that tests for selecting to
secondary schools should, under the Authority's admission arrangements,
take place after rather than before all parents had expressed
a preference for the school they wanted their child to attend.
The implications of the Torbay Determination are that tests for
selection to secondary schools should follow and not precede the
invitation to all parents to express a preference.
4. Management of the problem in areas that are not wholly selective
but have wholly selective schools within their area
4.1 LEA practice:
A system adopted by a number of areas
which are not wholly selective but have one or two selective schools
in their area is to ask parents to express preferences for both
selective and non-selective schools on a common form without first
knowing if their child is eligible (i.e. has reached a high enough
score in the test) for a place at a selective school. The tension
mentioned above between the inclusive and exclusive principles
is addressed directly by some LEAs. For example one LEA states
clearly in its composite prospectus:
'If you enter your child for the 11plus you should
bear in mind that a place at your local comprehensive school or
community college cannot be reserved in the event that he or she
does not gain a grammar school place.'
In this case advice was then given about the previous
year's entry (what marks children had and what marks those who
were given a place had). This would have had the effect of reassuring
those parents whose children's measured attainment to date was
well above average but would put pressure on those considered
borderline. It represented a risk for these parents but it allowed
most parents to make a reasonably informed choice. Of course the
'pass' mark may change each year as the number of students achieving
those marks increases or decreases. The local authority acknowledged
the fact that,
' … A place is not guaranteed for a candidate
who qualifies for a place under the 11plus selection arrangements…The
guarantee of a grammar school place was necessary in the past
when the only alternative was a secondary modern school, which
did not cater for pupils of grammar school ability. Nowadays suitable
alternatives are available at comprehensive schools and community
colleges that cater for the whole academic ability range.'
This robust position is replicated in a number of
other areas and would tend to facilitate comprehensive (all-ability)
intakes at the non-selective schools in the area.
5. Access to tests for places at a selective school
5.1 Opting in
or opting out? A significant aspect of
the admission process to selective schools is how children are
entered for the test. We found that there were different ways
in which this happened. In some areas all children in the last
year of primary school were entered for the test (with parents
able to withdraw them by request). This method of entry was common
for wholly selective areas. It provides universal access and raises
no obvious equity issues beyond those generally associated with
selective systems. In other areas children were entered on the
primary school's recommendation (with parents able to include
them by request). This involves, as a formal part of the process,
a discussion with primary school staff about the child's secondary
education and raises questions about the status of primary schools'
judgements and recommendations. Because there is an element of
judgement at this early stage in the child's school career it
is possible that some primary schools may inappropriately close
off an opportunity for some children.
5.2 Possible problems
with non-universal systems: There is also
the possibility that parents will differ in how they act on the
primary school's recommendation. In addition we know from the
extensive work on how parents choose a school that there are significant
differences in the way that parents from different social groups
respond to information and to the task of choosing. In other LEAs,
and for all wholly selective Voluntary Aided and Foundation schools,
parents must request that their child take the test as part of
the application to the school. Skilled and semi-skilled choosers,
who tend to be from more advantaged families, will do this more
readily than those who are less engaged with the process of choice.
6. Evidence as to the educational benefits of
selection
6.1 Evidence from
Schagen and Schagen and PISA: Members asked about the evidence
as to the educational benefits of selection. I explained that
as I read the evidence the differences between the two systems
in terms of exam performance was very small. I quoted two pieces
of recent evidence and present them here with more precision than
was possible in the hearing. Ian and Sandie Schagen of the NFER
have produced authoritative work (Schagen and Schagen 2002) concerning
the effects of selection in 149 LEAs in England. They considered
the effects at the level of the school (i.e. for the pupils attending
grammar or comprehensive schools) and at the LEA level (the effects
for the whole cohort of children in the LEA). At the level of
the school there were significant differences but at the level
of the LEA there was no significant difference between selective
and non-selective. Specifically they concluded that at the level
of the school 'there is a large 'grammar school effect' which
shows that borderline pupils - those who narrowly obtain a grammar
school place - obtain much better GCSE results five years later
than pupils of equal prior attainment in comprehensive schools'
and on the other hand that 'pupils of higher prior attainment
(key stage 2 average level greater than 5)' fared better in
comprehensive schools. They hypothesise that these two gains at
school level balance out so that there is no difference between
the systems in the performance of children at GCSE at the LEA
level.
The PISA study looked at the different performance
of national systems (OECD/UNESCO-UIS (2003). We should be very
cautious about cross national comparisons but the findings are
of interest. This study was concerned to investigate any relationship
between school and student characteristics and student performance.
They found that 'the impact [on educational performance]
of school's socio-economic background is much stronger than the
effect of any other variable (including school climate, homework,
reading engagement etc…..) Ch. 7 p219. They also investigated
the factors that were associated with social selection and found
that academic selection is correlated with socio-economic segregation,
'Among the ten countries with the most pronounced socio-economic
segregation observed in PISA, all carry out selection procedures
that channel students into different streams of secondary education
before or at the age of assessment.' Ch 7 p220.
7. Proposals and recommendations
7.1 The need to
accommodate different solutions: Any set of recommendations
needs to be sensitive to the fact that different areas experience
different levels of problems. Policies designed to solve chronic
problems in London should not make things worse in Cumbria.
7.2 Issues to be
considered: We know that the majority of parents want their
children to go to good local schools, defined not just in terms
of exam performance but also in terms of moral and physical security.
They also want predictability and speediness in the admission
procedures. The evidence suggests that balanced intakes will reduce
the problems of polarisation and are either educationally neutral
or beneficial. Further, local contexts must be taken into account.
It will therefore take a combination of measures to affect these
connected issues. Some proposals are given below.
7.3 Proposals and
recommendations
Proposal One: The aim of amendments to admissions
policies should be to:
· encourage
balanced school intakes;
· maximise
parental satisfaction with the process, but more importantly,
the outcome;
· increase
social justice and maximise overall choice (i.e. not give choice
to some by taking it away from others);
· facilitate
cooperation between schools;
· maximise
efficiency and cost-effectiveness as far as possible consistent
with other principles;
· have
concern for wider environmental implications such as traffic congestion.
Proposal Two: Responsibilities for admissions
to schools should rest with one admission authority for a given
area (i.e. VA and Foundation schools should no longer act as the
admission authority for their school). One
of the things that increases segregation of intakes and increases
the difficulty of managing the process of admissions for both
parents and admission authorities is the existence of a number
(sometimes a majority) of autonomous admission authorities in
an area. The new Code (DfES 2003) has made significant moves to
increase collaboration and cooperation but it stops short of taking
away the ability of schools to act as their own admission authority.
There should be consideration of combining the admission responsibilities
into one authority for a relevant area. This seems particularly
important in London. Clearly there would need to be extensive
discussion with interested parties but the relevant area for London
would most effectively be drawn to include the whole area where
presently there is a great deal of cross border admission. Unifying
admission authorities would greatly improve the appeals procedures
of Voluntary Aided and Foundation schools which are at the moment
poor (Coldron et al 2002).
Proposal Three: Admission criteria should be common
within the area covered by an admission authority.
A single authority for an area implies that admission criteria
should also be common for community, Foundation, Academies and
CTCs. For religious schools there would need to be an additional
criterion as to religious commitment common to all religious schools
in the admission authority but in combination with ability banding
to prevent covert selection.
Proposal Four: Admission authorities should seek
to maximise parental preference overall rather than giving absolute
priority to the first in a ranked list of schools. This
is already a recommended model in the new Code of Practice. It
allows admission authorities to balance other criteria such as
proximity with parental preference to gain optimal satisfaction.
Proposal Five: Schools should be encouraged to
join into collaborative federations and there should be active
exploration of the possibility of making federations the unit
for admission rather than the individual school. This
could contribute significantly to reducing polarisation of perception
especially if Proposal Eight underpins it and banding by attainment
is used to balance intake.
Proposal Six: The major oversubscription criterion
for community and Foundation schools should be proximity (or catchment
areas based on proximity) and, for Voluntary Aided schools, catchment
areas. This is the least worst option
and would have the undesirable effect of reinforcing selection
by mortgage. Therefore this Proposal needs to work in conjunction
with Proposal Eight on resourcing harder to educate children and,
in urban areas, the use of banding (perhaps within groups of schools)
to mitigate segregation of intake. Admission authorities need
to ensure that all schools are able to accept casual admissions
of harder to educate children.
Proposal Seven: Selection by general ability should
be phased out. This includes partially
and wholly selective systems. They are not radically more or less
effective educationally but they create greater segregation, generate
a lot of appeals and are incompatible with parental preference
systems.
Proposal Eight: Resources to schools should follow
the child and be based on a calculation of risk factors indicating
how easy or hard each child is to educate. A school with a greater
proportion of harder to educate children will receive proportionately
more resources than a school with fewer children who are harder
to educate. We know certain characteristics
of children (such as the socio-economic status of their family,
their prior attainment, their mobility between schools, and their
first language) are highly correlated with their educational attainment
at 16. Some are easier to educate than others. We also know that
some schools attempt to select the easier to educate and to exclude
the harder. The extra resources would encourage schools to better
balance their intakes and this would in turn reduce one of the
pressures toward segregation and polarisation of perception. It
may also have other advantages; for example it would go some way
to complement the advantages already experienced by pupils in
schools with high socio-economic status intake; it encourages
an inclusive rather than an exclusive response from schools; it
acknowledges the additional difficulties that schools as institutions
have when they have educationally disadvantaged intakes; it would
help to reduce some of the competition between schools and therefore
provide a sound foundation for the operation of the Admission
Forums or the unitary admission authorities and the development
of Federations of schools; it addresses one of the main problems
of English education which is the low achievement of the middle
and lower attainers; it would improve the morale and the retention
of teachers in the schools most in need of stability.
Proposal Nine: Greater travel subsidies should
only be considered as one among a number of ingredients. Greater
real choice may have a role in reducing segregation of intakes
in some areas but should not be considered as either adequate
in itself or as problem free e.g. an increase in the school run
traffic would be an undesirable result. It should be considered
as an option in specific contexts.
References
Coldron, J., Stephenson, K., Williams, J., Shipton,
L., Demack, S. (2002) Admission Appeal Panels: Research Study
into the Operation of Appeal Panels, Use of the Code of Practice
and Training for Panel Members School of Education, Sheffield
Hallam University for the DfES, Research report RR344
DfES (2003) School Admissions Code of Practice
London: Department for Education and Employment
Flatley, J., Williams J., Coldron J., Connolly H.,
Higgins V., Logie A., Smith N, Stephenson K. (2001). "Parents'
Experience of the Process of Choosing a Secondary School."
Office for National Statistics and Sheffield Hallam University.
DfES Research Report RR278
OECD/UNESCO-UIS (2003) Literacy Skills for the
World of tomorrow - Further results from PISA 2000 OECD
Schagen, I. and Schagen S. (2002) Using national
value-added datasets to explore the effects of school diversity
presented at BERA, Exeter, 12 September 2002
Williams, J., Coldron J., Stephenson K., Logie A.,
Smith N. (2001). "An analysis of policies and practices of
LEA admission authorities in England." in Parent's Experience
of the Process of Choosing a Secondary School, edited by J.
Williams and J. Flatley: Department for Education and Skills
| 
