Improving test quality: A teacher’s role in it

Anything that is out of the ordinary or the norm requires some extra bit of work. Teachers at the outset need to start by designing the test specification in line with the intended objectives of the course

A general and a layman’s understanding of the term ‘test’ is that students take it and teachers give it, with the former deemed as having understood the taught concepts after managing to pass the test. More often than not, it is the test takers’ fault if he/she fails the test. On the other hand, it is the examinees again who get the credit, sometimes even undue, for what might be called ‘passing with flying colours’.

The test papers and items, by and large, do not get much scrutiny.  However, the effectiveness of the test in actually measuring what it purports to measure is also of paramount importance to pedagogy with poorly devised tests giving a false picture of the student’s true caliber. Teachers, who also double as test constructors, can greatly contribute to increasing the quality of tests they give by adopting a few tested and proven measures in the pedagogical field.

In the context of Nepal’s schools, standardised final level examinations are given at the end of the basic level in grade 8, and in grades 10, 11 and 12 at the secondary level. The curriculum has been revised recently, and many in the field hope that it will address the problem of 10th grade pass-outs as the high school level course was found to be too difficult in the past.

Thus the need to improve the quality of tests and the corresponding curriculum to bridge this gap was always there.

This is called ‘predictive validity’, which is a degree to which a test can predict candidates’ future performance. This much-needed change in the quality of such standardised test papers is cardinal to school teachers as they are the ones who prepare the students for such tests. However, since test construction here is out of their jurisdiction, they can only hope that such a need is felt by people concerned at the macro level as well. In the meantime, what they can indeed do is try and make the quality of their internal tests better.

Rote learning by students and teachers’ teaching students formulaic responses to oft repeated questions have become a general feature in grades 8 and 10, for the purposes of the Basic Education Examinations (BEE) and Secondary Education Examinations (SEE) respectively. If one were to adjudge the quality of these question papers against the designated test specification, there are hardly any discrepancies; the test is slavishly faithful to the specification.

However, the specification itself leaves a lot to be desired as it is a disproportionate mix of predominantly knowledge level, or fact recall, questions. This might be the reason why the teachers then simply resort to ‘teach to the test’ tactics. Why bother about challenging the students with a higher order of thinking if ultimately it is only the ‘who’ ‘what’ ‘where’ questions or the mechanical ability of sentence-type transformation that matters? Such, unfortunately, is the test impact of teaching and learning in the classroom, which is also known as the ‘washback’ or ‘backwash’ of testing.

Against this backdrop, this writer believes, the quality of internal tests becomes particularly important. If constructing tests at variegated knowledge and skill levels is felt to be too risky and superfluous by teachers at these levels, they can still be done at other levels.

Anything that is out of the ordinary or the norm requires some extra bit of work. Teachers at the outset need to start by designing the test specification in line with the intended objectives of the course.

However, for reasons of convenience, what the teachers usually do is make slight adjustments to past question papers or randomly select questions from the chapters which they have covered.

Test specification serves as a blueprint for teachers both in terms of preparing the students for the test, having a proper understanding of what they need to learn as well as the

eventual construction of test papers. A good test specification considers many important factors like setting out clear objectives, timing, format, rubric for students as well as marking scheme for the examiners. It also needs to make sure that the content it stipulates for the test has diversity in the objectives with respect to addressing the various cognitive levels in students.

Bloom’s taxonomy, proposed by educational psychologist Benjamin Bloom in 1956, is often referred to for this purpose of ensuring richness and diversity. It is simply a classification of different objectives by educators at different levels of learning: remembering, understanding, applying, analysing, evaluating and creating. Apart from the first two objectives, the latter four, which actually help develop critical thinking in students, often get overlooked.

The advantage of making them a mandatory component of the specification table is that it surely results in the setting of the corresponding questions. The other and, perhaps, more significant outcome would be in classroom learning itself as addressing those objectives would require teachers to plan classroom activities in line with them.

Teachers, for well justified reasons, put the blame on the irrelevancy of the curriculum for not being able to deliver well. Granted, the curriculum leaves a lot to be desired, but within that curriculum itself, they still can go that extra-mile and do what is under their control. Thinking about improving the quality of our tests could be one way to start.

Neupane is with

Kathmandu University