9.1. The grade 7 examination
9.2. Assessing reading achievement nationally at grade 4
There would appear to be two main concerns with respect to the testing of reading in Zambian primary schools on a national scale. One is the nature of the Grade 7 primary examination; the other is the development of reading achievement tests at Grade 4 as part of a national assessment system.
9.1.1. It is widely recognised that nationally set school examinations have a powerful washback effect upon teaching. As far as Zambian primary schools are concerned the Grade 7 examinations are especially important as they control access to Junior Secondary School. While an exhaustive analysis of the examination would be out of place here, it is worth briefly considering the reading component of the examination in order to assess its possible influence on teaching, and how it might be modified to enhance the quality of the teaching of reading comprehension. The following comments are based on a consideration of two papers, namely the 1988 English paper (S/YE1/G7/88), and the 1990 English paper (S/PE1/G7/90).
9.1.2. The examination consists of 3 parts. Part 1 examines grammar, and Part 2 punctuation and inter-sentential cohesion (in effect, an indirect attempt to assess writing skills). Part 3 tests reading comprehension through three passages. The shortest passage is around 250 words, the longest around 500, with the other four close to 350 words. Each passage is followed by 6 comprehension questions. These questions, like all others in this examination are multiple choice (4 options in each question).
9.1.3. The reading comprehension component of the examination will be dealt with here in terms of the following: first the language level, second the text types, third the multiple choice format, and fourth the reading skills.
(i) Language level: this seems to be relatively high bearing in mind the weak overall performance of Grade 6 pupils in the Word Find test. It is difficult to see how most Grade 6 pupils can in the space of one year come up to a standard where they would be able to perform reasonably on the Grade 7 English examination. The problem here is that pupils of average ability have little chance to show what they can do, but only what they cannot do. This presumably derives from the examination's primary function as a selection device for Grade 8. Confining the examination to this function means that it only needs to discriminate at the top end of ability.(ii) The text types: the text types in the two examination papers looked at were continuous prose texts consisting of narrative and simple exposition. While these are perfectly acceptable as far as they go, there is room for the inclusion of a greater variety of texts which could be more relevant to the "real-life" situation of the majority of pupils (eg forms, instructions, advertisements etc). There is also room for considering text types used in the other subject areas of the primary curriculum (which are arguably also "real life" texts for primary school children). A further advantage of extending the range of text types in this way is that they would be susceptible to more realistic test tasks.
(iii) The multiple choice format: the examination relies exclusively on the multiple choice format. The problems arising from this are:
(a) the distractors presented can do their job too well and confuse some pupils who might otherwise have understood the point at issue.(b) the need for distractors increases the length of the paper and the net result is a relatively low number of questions
(c) it is possible that some fortunate pupils can do well through guessing
(d) copying is relatively easy with multiple choice
(e) the backwash effect of this format seems to be that the testing of reading comprehension in lower grades is confined to the multiple choice format
(f) the format is artificial and can only awkwardly be adapted to reading operations such as those mentioned in paragraph (iv) below.
The overwhelming advantage of multiple choice for large examinations is that it lends itself to machine marking. However, against this should be weighed the above limitations and its lack of sensitivity as a testing format. It directs pupils towards limited choices, which they might not have thought of in a free reading situation. Consideration should therefore be given in the reading comprehension component to including short written answers ie one word, one phrase or single sentences, which would make for assessment more in line with current educational objectives.
(iv) Reading skills: while opinion is divided as to the exact nature of reading skills, such higher order skills as identifying main points, identifying supporting detail, making appropriate inferences, or interpreting material are not easily addressed with the multiple choice format. A move towards incorporating a range of reading skills would also be in keeping with the new aims for Grade 7 outlined in the syllabus for the new ZBEC (see 6.2.2), and could also encourage abilities needed to learn through reading in other areas of the primary curriculum. (These higher order skills, of course, presuppose a competence in basic literacy, which should not be the concern of the Grade 7 examination).
9.1.4. In view of the above it would seem worthwhile to investigate the possibility of modifying the current Grade 7 English comprehension section, with attention given to:
(a) extending the range of ability assessed, in particular towards the middle ability bands. In principle it would be possible to do this without sacrificing the discriminatory power of the examination at the top end.(b) extending the range of text types (or genres to use a current term) so as to encompass more "real life" texts, including text types relevant to other primary subject areas.
(c) extending the range of reading skills tested. This might necessitate the use of formats other than multiple choice. There would be implications here for the marking process, but the costs of such an operation would have to be weighed against the beneficial effect such an innovation could have on the teaching process.
9.1.5. I would accordingly recommend that specialist research be carried out into:
(a) specifying the reading comprehension abilities relevant to real-life situations, relevant to reading to learn in other subjects in school, and relevant to the needs of the majority of pupils who will end their schooling at Grade 7. This would involve, among other things, a consideration of appropriate topics, and text types for such pupils. It should also consider relevant reading skills and degrees of fluency in reading. Such work should be carried out in conjunction with the Curriculum Development Centre, and be integrated with the aims of the new ZBEC syllabus (see the extract from the new draft syllabus in Appendix M).(b) investigating examination formats which lend themselves to assessment of the selected reading abilities. It is likely that a compromise between practicality and validity would be needed here. Practicality involves issues such as the length of the examination paper, and the resources available for production and marking. Validity refers to the degree to which the test reflects an theoretically acceptable approach to reading.
Clearly to leave the current reading comprehension component of the Grade 7 examination unmodified could undermine a more comprehensive approach to reading as well as the aims of the new syllabus.
Following the Jomtien Conference Zambia formulated its own goals and strategies for the 1990's. A part of this involves the assessing of learner achievement through competency tests for diagnostic use at Grade 4 (see Kelly, 1993). In advance of the main project, Zambian educationalists intend to prepare and administer their own interim tests in, among other areas, reading. The tests are to be administered annually to a sample of 1.5% of Grade 4 enrolment, with a view to obtaining information for a proposed National Assessment System. This proposal is to be welcomed, and one to which the present project might serve in a contributory capacity. As the proposal is already going ahead, it would be presumptuous to make recommendations other than a general recommendation that the project be supported.
