Research Papers
Can Testing Really Raise Educational Standards?
Harry Torrance
(Institute of Education, Manchester Metropolitan University, UK)
Professorial Lecture delivered at the University of Sussex, 11 June 2002
Introduction
It's a strange Alice-in-Wonderland World we enter, when we enter the world of examinations and testing; exam passes and test scores have never been higher, as I will demonstrate later in this lecture, and yet the moral panic surrounding exams and educational standards continues unabated, as a glance a any recent education headline will demonstrate. Interestingly enough however current newsworthy stories have a slightly different angle to them, compared to the usual 'shock, horror, standards are falling' headlines. Suddenly the headlines are beginning to hint that we may have too much testing, rather than too little, with the system buckling under the weight of the current load.
Exam Boards have so many papers to process, and results to report, that serious errors are creeping into their procedures - Edexcel is hardly out of the news these days, for faults in their question-setting, while the Scottish Qualifications Authority had that debacle over reporting thousands of wrong results last year. Meanwhile in the Key Stage tests Headteachers are under so much pressure to deliver results that some resort to cheating, while students and teachers alike complain of the exam overload brought about by Curriculum 2000 and the introduction of A/S levels. To use an engineering metaphor, it seems that we are beginning to 'test the system to destruction'. Well, that's all very well when we want to know how much force the materials in a bridge can withstand, but it hardly seems appropriate to the future building blocks of our society - our children.
So what on earth is going on here? How did we get into this state and why as a society are we so obsessed with testing and test results? Using testing to improve schooling seems compellingly simple - set your goals, measure whether or not they have been met, do something about it if they haven't. Surely testing can indeed raise educational standards? But if so, how, and with what consequences? And if not, how else should we assess and develop our children and our schools?
The UK is not alone in turning to assessment to reform and control schooling. Governments around the world have become concerned about educational standards and their implications for economic competitiveness (Torrance 1995, 1997). Likewise policymakers and educationists around the world are interested in using assessment to try to raise standards; but only in the UK, or, to be more accurate, England, has the focus become so narrow, intense and, as I will argue in this lecture, counter-productive.
In this lecture I shall be looking at:
-
Government approaches to setting and measuring standards over the last twenty years;
-
The logic of using testing to raise standards;
-
The impact of national testing on standards;
-
The scale and costs of the enterprise;
-
What is to be done? Alternative approaches to testing, and the development of formative classroom assessment.
I will have to skim over some elements of the argument pretty quickly, but it is important to document and reflect on where this impetus for testing has come from, before moving on to look at current impact and future prospects. I shall focus mainly on the UK but bring in research evidence from elsewhere, particularly the United States, as appropriate.
Please click on the link below to download a pdf file of this paper:
Can Testing Really Raise Educational Standards? (pdf format, size 435KB)
