Examination of Witness (Questions 240
- 259)
MONDAY 7 APRIL 2003
PROFESSOR CAROL
TAYLOR FITZ-GIBBON
240. The average points score would be the key,
would it?
(Professor Fitz-Gibbon) The average points score means
that every student counts so the behaviour implied by using the
average points score is to look at the progress of every child
and that is the kind of ethics that teachers want. It is unethical
to care more about the D student than about anybody else.
241. But a school that only puts children in
for easier subjects, even that system could be defeated?
(Professor Fitz-Gibbon) A C in anything as long as
it is a C. This is not good because it is a very, very bad indicator.
Let me give you a comment I heard a student saying to another
student: "Do you want help? Tell her you are a C/D borderline",
so even the children know if you are a C/D borderline you will
get extra help. I mentioned this at a meeting of researchers with
someone from the DfES and the man from the DfES was the only one
not amused. He said that Ministers would be very annoyed if they
thought schools were concentrating on D students. I was taken
aback and he said, "You are only giving us an anecdote",
so I dredged up the meetings I had been at where they were giving
money for D students. Over coffee the other researchers said to
me, "What schools does he know about? Where does he live?
Show me a school that is not concentrating on D students."
However, let me say this: they did not say that to the man who
worked with the Minister because universities are looking for
contracts, so they do not speak up, they are not going to be difficult.
I thank schools enormously for supporting the staff so I do not
have to think will someone lose their job if we do not get the
next contract and I can speak without fear or favour. Until universities
are soundly funded they will keep their mouths shut, which is
terrible.
Valerie Davey
242. What account does YELLIS take of youngsters
for whom English is not their first language?
(Professor Fitz-Gibbon) It is not meant
to be a label on a student. It is not an IQ label. To get away
from that we produced chances graphs that show from any score
they can get any result, they just have different chances. The
fact that English is their second language will depress their
vocabulary score but it also makes it harder for the teacher to
teach them, that is the point about the YELLIS baseline or the
MIDYIS baseline or the ALIS baseline. How hard is it going to
be for the teacher to get the value-added progress? It does not
matter that their score is depressed, but it does make it harder
for them to get through the English and possibly through the mathematics
if their English is really bad.
243. I happen to believe that children who have
two languages are doubly able. I would hope that they would score
very high.
(Professor Fitz-Gibbon) I think we should have nursery
schools where you decide what you want your child to be bilingual
in. Cantonese? Okay, there is a Cantonese nursery school there.
Spanish? Okay, there is a Spanish nursery school there. They absorb
languages, there is a module in the brain that learns languages
with no effort at all. You cannot stop them learning so why try
and teach it later? Put them in a nursery school and get a bilingual
child.
244. What does YELLIS do in terms of children
with at least two, in some cases three, languages? Do you know
about that in that test or does the test overcome that or does
it have no relevance to that?
(Professor Fitz-Gibbon) We collect the data and that
is recorded. In the MIDYIS test, which is the baseline for all
the secondary schools, and is more complicated than the YELLIS
test, we will report a whole profile of separate scores, so we
are aware of that and we will point that out in the reports.
Valerie Davey: I still have not got an
answer to—and let's stick with YELLIS which is the one we
have concentrated on today—is the YELLIS test more difficult
or less difficult or equally difficult for children with a background
of two or more languages? In other words, is it language sensitive?
Chairman
245. Bangladeshi students, kids from a Bangladeshi
background, from Pakistan?
(Professor Fitz-Gibbon) It depends entirely
how long they have been here and what their aptitude for picking
up languages is.
Valerie Davey
246. But it is vocabulary based?
(Professor Fitz-Gibbon) There is a large
vocabulary section for the very good reason that vocabulary and
maths are the biggest predictors in success at GCSE. That is not
to label the child. In fact in the early years of the test I would
not give the results to the schools. I would say, "This is
just for us to tell you the value-added later." It is a baseline
against which we can measure how difficult it is for the school
to get them through the next exam.
247. So you should not know from the outcome
whether a child has one language or two? They take the test and
it is irrelevant to you and irrelevant to the school?
(Professor Fitz-Gibbon) It is not irrelevant but what
do you suppose we should do?
248. I do not think you should know. I think
there are some very bright children with two or three languages
who will do supremely well in any test and there are others who
will not. My hope was that you would say it is irrelevant but
you are not saying it is irrelevant.
(Professor Fitz-Gibbon) We have a question about is
English an additional language, if so, what else is spoken in
the home? It varies from one project to another so the exact wording
I do not have at my finger tips. That is so we can subsequently
do research and answer questions that people will ask us on how
do people with EAL perform under the YELLIS test.
249. How do they perform?
(Professor Fitz-Gibbon) Their vocabulary of course
will generally be lower but it depends on how recently they have
come. All of these things are generalisations of completely overlapping
distributions so you have to look at the individual child.
250. If you look at the individual child in
that respect, why do you not look at the individual child in terms
of their background, in terms of how many books they have got
in the house, all the other things we have been talking about
this afternoon?
(Professor Fitz-Gibbon) We ask exactly that question,
it is one of our cultural capital questions. How many books in
the home? Does anyone ask you about homework? How many times have
your parents been in the school? A lot of questions like that.
That is all there but the predictive validity of that does not
come anywhere close to the predictive validity of YELLIS. It is
a baseline value-added about that general aptitude that any teacher
recognises is a good predictor of subsequent achievement and is
the baseline that is fairest for teachers.
251. Who answers all those questions, the child?
(Professor Fitz-Gibbon) The child in the confidential
questionnaire that they receive in a plastic envelope and it comes
back to the university. For administration a tape recording says,
"Do not raise your hand if you have a problem with the questionnaire
because the teacher will not look at your answers." It is
there on the tape recording. We go to the trouble of sending a
tape recording so in every school the test is administered in
the same way with same words.
252. I am sorry I am obviously so thick this
afternoon but are these questions part of the 25-minute test or
is this background information before they do the test?
(Professor Fitz-Gibbon) After they have done the test,
it is the questionnaire.
253. There are three different things, sort
of unrelated, but covering across the matters we were talking
about, the first one is that you talked about the problem of the
league tables and the five As-Cs, and so forth, distorting the
way that children are advised to go in to maths or not go in to
maths because it is hard. There are all sorts of complicating
factors there as well. I used the YELLIS system for 10 years as
a head of sixth form, and extremely good it is, however even then
counselling GCSE students after they got their results in the
summer, and you would have a student who got seven grade Cs who
would say, "I want to do maths, physics and chemistry at
A-level". Using your chances graph from ALIS you would be
advising them that they would be better doing other things. We
were not doing it because we wanted to manipulate the school's
scores, because we had a head who said we should not be doing
that, quite rightly, we were doing it for the best career prospects
of the student, but it would lead to the same distorting effect.
By using your method you were arriving at the same distorting
effects as the league table emphasis on five As-Cs. How can you
get round that?
(Professor Fitz-Gibbon) I think the school has to
take into account the fact that maths, science and a foreign language
A-level will lead to better employment prospects than sociology,
and psychology even.
254. Not if the student is likely to struggle
and fail.
(Professor Fitz-Gibbon) But they do not from a C,
the chances graph do not show they will fail. In fact that is
what made me start the chances graphs because I knew that some
colleges would not let you do A-level maths if you only had a
C. I thought, I wonder what the chances are, let us look at the
other colleges and see what happens to those who go in with a
C for A-level, and sure enough, it is like the 11 Plus, to have
a sharp cut is absolutely wrong because people have different
rates of development, they are on-line, they are off-line, so
chances graphs have moved us away from a single predicted grade
to show a spectrum of possibilities. I do think there is certainly
an incentive to schools to keep people out of difficult A-levels
because if the current method of value-added calculation continues
it is absolutely ridiculous, and they are going to lose out by
putting people in for good subjects—except that a bright
student is more secure doing A-level maths than A-level English
because of the accuracy of the marking in maths compared with
English.
255. The other point about league tables is
the fear, the opinion that they lead to pressures on schools,
they lead to unfair comparisons because with a league table you
are comparing very different types of schools which say you should
all be doing the same and achieving the same. What some heads
would say is if they have a school with a low socioeconomic intake,
like in an inner city school, the league tables are going to be
unfair to them. You are saying on the one hand socioeconomic factors
do not make a difference, although on the other hand concentrations
of them in segregated schools do make a difference. Is that right?
(Professor Fitz-Gibbon) Yes. If you are going to have
value-added tables it should be based on prior aptitude or prior
achievement, or both. I share the concerns that a number of you
expressed about reporting the aptitude back to the school in case
it cuts off possibilities for students. As I said, in the first
few years I would not give them the results but then I felt I
must trust the professionalism in schools and let them decide
how to use the data. We talk about distributed research, we are
a research arm for schools and the power is in both corners, as
it were. From the data we can discuss the way forward. When schools
are under pressure, and I had not realised this, they will shift
the entries, or the students will choose to get a higher grade
and will not be counselled otherwise e.g. "Maybe you should
have a go at maths if you want to have a go at it."
256. You say the problem with poverty is not
that it causes low achievement, and you then say, because schools
are often segregated by social class it does look as if poverty
is a powerful factor. Are you making the distinction that individuals
may well have the ability to achieve and their background will
not matter from that point of view, but if they are all concentrated
in an inner city ghetto school or a sink school then collectively
the weight of the social background of the pupils will make a
difference?
(Professor Fitz-Gibbon) Actually that is called the
compositional effect, does the average ability of a group in the
classroom have an influence? You can imagine that it might, if
everyone is doing their homework you perhaps would tend to do
your homework and if nobody is you would feel a fool doing your
homework, would you not. It may have an impact in secondary school
in particular, when the peer group is so powerful, and far more
powerful than parents. We do not know that for sure. We need to
watch that very carefully. The reason not to segregate is that
it is unfair, it is not a fair distribution so that everyone has
a similar chance. The most important thing is to have really good
teachers. The large variation is attributable to which classroom
they are in. We have reported on a value-added national project,
34% in English, 40% of the explained variance in mathematics,
for which classroom they are in, and in the school it is only
a matter of 5% or 10%. It matters that we have happy teachers,
and happy teachers should be in control of what they are doing
and should not be subject to things that would be illegal in the
United States, like Ofsted.
Chairman
257. I wondered when you were going to mention
Ofsted!
(Professor Fitz-Gibbon) I did not mean
to mention it, it just came up.
Chairman: I never mentioned Chris Woodhead.
It is interesting both of us have this thought.
Paul Holmes
258. There was an article in the Educational
Press a few weeks ago from a governor of the Riding School
in Halifax, the one which notoriously hit the headlines some years
ago, he had been a governor all of the way through before, during
and after and he said, "It got a really bad reputation, it
hit the headlines, it had the super heads brought in, the extra
money, the wonderful teachers, all the attention", and eight,
or however many, years later it is still not doing that much better
in terms of GCSE scores than it was at the start, even though
it had all the razmataz and the wonderful new teachers. He was
very much making the case there that because there was a segregated
system of schools selecting left, right and centre, and the Ridings
was right at the bottom of the pecking order, that it therefore
got a much higher level of socioeconomic deprivation amongst its
pupils, and that critical mass meant that however good the teachers
were they still were not going to overcome that problem.
(Professor Fitz-Gibbon) Intuitively I
feel that that is very likely the case. I remember when somebody
on the road did not get through the 11-Plus, they put on a different
uniform and went to the secondary modern school instead of the
grammar school, it was a terrible thing for that child. We label
them with uniforms: they go to a so-called good school. We should
stop all this nonsense about good schools and bad schools. Which
year are you looking at? The value-added goes up and down like
the weather, so it is good one year, bad another. When they are
at the bottom they are going to go up, regression to the mean.
A super head may come in but regression to the mean more or less
promises it will go up. It is not the school, although you know
a malfunctioning school for some particular reason would be very
rare, it is the teachers that matter, and if you get a school
with a bad reputation so nobody wants to teach there you will
not get the best teachers in the classroom. It is the whole system
where we need really good teachers because they are appreciated
and they are dealt with fairly on objective data.
259. You say at the end of your submission that,
yes, there is more money going into the education system with
lots of ideas on how to improve it but it is not really based
on evidence and you are very critical of the role of special advisers
who are not experts, do you want to elaborate on that?
(Professor Fitz-Gibbon) No!
|
