Latest:

13:12 Tank marks charity eggstravaganza
13:12 RBS bankers arrested in tax probe
12:12 Sex change man 'first male mother'
12:02 BAE executives 'could net bonuses'
11:52 Taylor urges FA action over Suarez
11:32 Hunt backs FA over Terry decision
10:52 Mancini open to Tevez return
10:02 Stars 'sing Whitney into heaven'

Professor Tony Gallagher, head of the School of Education at Queen’s, answers readers’ queries....

Monday, 31 August 2009

Why can’t grammar schools accept either test results?

Q Is there any reason why all grammar schools should not accept either the GL test results or the AQE test results for consideration for entry to their school?

While scores cannot be compared directly, percentiles can. So if a child comes in the top, say, 25% on either test, he or she has achieved what was formerly an A grade. Similarly for further percentiles he or she would have received a B1,B2, C, etc.

Perhaps the only reason for not accepting this way of reading across the two tests is if either test is faulty or biased and does not accurately rank the top 25% etc, but muddles them up.

—from Michael Dubitante.

A In the 11-plus, the top 25% of pupils were awarded an A grade, the next 5% were awarded a B1 grade, the next 5% a B2 grade, and so on. The percentages were determined by the total number of P7 pupils: thus, if only half of all P7 pupils took the tests, then half of these pupils would be awarded an A grade. This assumed that any pupils who opted out would have achieved a D grade. This assumption, accurate or not, could be made because only one test operated.

Since pupils self-select for the GL or AQE test systems, it would be inappropriate to make any assumption about the potential performance of pupils who do not take the tests. Similarly, it would be invalid to compare the results in the way suggested.

This creates a dilemma for grammar schools prepared to accept GL or AQE results as there does not appear to be a valid basis for accepting a pupil with, for example, a GL test score while rejecting a pupil with an AQE test score, or vice-versa.

Most grammar schools seem to have anticipated this by declaring that they will consider the results of either the GL or the AQE tests, but not both. Most grammar schools seem to have anticipated this by declaring that they will consider the results of either the GL or the AQE tests, but not both.

Email questions to: ktorney@belfasttelegraph.co.uk

Print
Text Size
- Normal
- Large
- Extra Large

Despite asking "It is a statistical question can I have an expert answer please?" Michael Dubitante seems to be answering his own questions on testing and with some inside knowledge. He neglects to advise parents that GL Assessment is the company that provided the CCEA 11-plus before a breakup. Perhaps he will now go into some detail about the methodology to be used to determine equivalence in the AQE/CEA/GL Assessment tests. Will it be a comparison of results from the norm – reference method of standard setting or the test-retest and inter-rater reliability of the modified Angoff method.
The game playing by insiders is only further muddling the minds of parents interested in equality of opportunity for their children.

Posted by Parental Alliance for Choice in Education | 04.09.09, 08:09 GMT

There will be a method of finding an equivalence between the two tests. There are differences in the construction of the tests, the AQE is developed in Northern Ireland but has not been trialled on pupils, instead they have relied on a panel of teachers examining individual questions for bias (gender, culture etc). GL have provided an off the peg test which has been trialled and out of any 10 questions many would be thrown out in the process. There are differences in format which may not be such a problem, but I have gone through the 24 CEA practice papers and the 8 recommended GL multiple choice papers and one does seem harder to me, than the other.
Not withstanding this as long as they are both valid and discriminating and much the same pupils apply as hitherto, then I believe yes of course pupils can be rank ordered irrespective of which test they have taken. It is more an administrative task than a theoretical problem.

Posted by Michael Dubitante | 03.09.09, 10:44 GMT

Exasperated - your metaphor distraction fails to mention the name of the captain of the education ship of fools.
Parents are still waiting for an answer to the question from Mr Michael Dubitante. Can you help or are you part of the problem?

Posted by Parental Alliance for Choice in Education | 03.09.09, 06:57 GMT

There really is no need for any school to set a test for admission. There aren’t different types of children who can only be taught in a particular type of school: this is accepted by the 'grammar' schools as they have been admitting children with any grade (including a D in the old 11+).

Parents should send their children to their local school and insist it is a centre of excellence; challenge the work, marking and teaching provided. The focus on testing distracts from the real issues of quality teaching and learning: not all grammars are good, not all secondaries bad. Our system (composed of both grammar and secondary) is flawed and the system is underperforming; you can't cherry-pick bits and ignore the rest.

Consider, to revisit an earlier metaphor, the first class on the Titanic who thought everything was OK till the entire ship sank. (I can just hear the ghostly voices murmuring "Why do we need more lifeboats, sure I have a seat already!")

Posted by Exasperated | 02.09.09, 12:55 GMT

There is an inequality of opportunity created by having two tests administered along sectarian lines. The headmaster of my local state grammar school expects fewer Catholics to apply next year because the GL test creates two barriers. One is that the students have to take more exams if they apply to his school nas well as the local Catholic grammar, the second is that the Catholic primary schools will have less time to spend practising for the AQE exam. Up until now a substantial proportion of his intake has been Catholic.

This strength of this disadvantage does not work the other way around in practice. There are very few
Protestants at the local Catholic grammar despite it being much easier to get into.

There really is a need for the Department of Education here to give guidance on some sort of equivalence between the two tests to avoid this sectarian discrimination they appear inadvertently to have caused.

Posted by Michael Dubitante | 02.09.09, 10:27 GMT

There is an inequality of opportunity created by having two tests administered along sectarian lines. The headmaster of my local state grammar school expects fewer Catholics to apply next year because the GL test creates two barriers. One is that the students have to take more exams if they apply to his school nas well as the local Catholic grammar, the second is that the Catholic primary schools will have less time to spend practising for the AQE exam. Up until now a substantial proportion of his intake has been Catholic.

This disadvantage does not work the other way around. There are very few
Protestants at the local Catholic grammar despite it being much easier to get into.

There really is a need for the Department of Education here to give guidance on some sort of equivalence between the two tests to avoid this sectarian discrimination they appear inadvertently to have caused.

Posted by Michael Dubitante | 01.09.09, 17:05 GMT

Michael Dubitante,

Professor Gallagher heads the School of Education at QUB. He has recourse to experts within his team such as the head of PGSE Mathematics, the teacher of teachers, to answer questions on mathematics and statistics. The Belfast Telegraph should know this also.Perhaps they will put your question on the validity and reliability of construct comparisons issue to someone who can provide an accurate answer. Good luck. Parents will watch and monitor with interest.

Posted by Parental Alliance for Choice in Education | 01.09.09, 07:38 GMT

Victoria College have told me that they will compare applicants on the basis of where they fall in the distribution of percentiles of either test. This is not to compare the papers but to assume that both tests are valid instruments to test ability and how well students do determines whether they get a place or not. It is not comparing apples with oranges, and is indeed the basis on which foreign applicants to the professions or university places are accepted or rejected. This is not an issue special to Northern Ireland.
Professor Gallagher suggests that an assumption is being made about the grades that non applicants might achieve, and that because the exam is voluntary it somehow makes invalid a comparison of the performance of students who take either exam. This is irrelevant.
It is a statistical question can I have an expert answer please.

Posted by Michael Dubitante | 31.08.09, 12:05 GMT

The lack of knowledge on testing from both questioner and the "expert" must shock parents seeking clarity on this imposed private deregulated system. Mr Dubitante wants to compare apples to oranges with the two tests and then further compounds the problem by removing information in suggesting use of the flawed CCEA instrument, letter grades.
The "expert" has makes it obvious when he forgets to remind Mr Dubitante that in the 11-plus an A grade was awarded to the top the top 25% of pupils eligible to take the test not just those who sat it. Non-entrants made it easier for their classmates to obtain a place in a grammar school.
So much for the expert answers.

Posted by Parental Alliance for Choice in Education | 31.08.09, 08:47 GMT