|
Selection of Reading Texts
The following points should be considered in selecting texts for reading test items:
Selection must be based on the test specification to ensure a representative sample of text types are selected to test as many sub-skills as possible according to the syllabus.
Length of the texts: This depends on factors such as proficiency level of students and the type of sub-skill/s to be focused. However, the text should allow us to prepare a sufficient number of questions. It is recommended to follow the syllabus for deciding the length of the texts too.
It is essential that measures are taken to ensure higher reliability. It is obvious that objective test items are used but this is not sufficient. If many passages are used rather than one, it will provide an accurate measurement of learner performance. But practicality should also be considered. The questions should not be related: give a fresh start in each question.
The passages must be interesting. However, they should not be overexciting to arouse emotions of candidates during the test, which may distract their attention. The texts must also not be biased: if a text on Buddhism is chosen, it may favor some candidates.
The texts should not be selected merely because it is easy to set questions from those passages.
Testing Techniques
The techniques the tester chooses must be strong enough to demonstrate reading behavior of candidates. However, these techniques should not be barriers to the candidates to demonstrate their true reading performances. For example, when a reading question requires the candidates to ‘read and write suitable answers to the given questions’, some of them will receive lower marks not because their reading skills are weak but merely because they cannot write: the technique interferes them to demonstrate their true reading ability. The techniques included in this book require minimal demand on student writing ability.
Some of the techniques which can be used in testing reading skills are discussed below:
Multiple Choice Questions (MCQs)
This is one of the traditional forms of testing reading skills. Since it is objective by nature, a higher degree of reliability can be expected. There are at least two major pitfalls in using MCQs. First, it is very difficult to construct good MCQs, especially in the case of distracters. Second, the students will guess the answers or copy them easily from those seated nearby.
Binary choice Items
This technique can be considered as a variety of MCQ, where the candidates are asked to select one answer out of two options, ‘Yes’ or ‘No’. In this case it has only one distracter and there is 50% probability of choosing the correct answer by chance. Though it has the advantage of ensuring higher reliability due to its objective type (and it is easy to set these questions), its frequent use cannot be justified.
Summary cloze type
The tester writes a summary of the reading text. In this summary, some gaps are left for completion. The students are not penalized for spelling errors. This type too is objective and ensures high reliability. It allows the tester to set a reasonable number of test items on a relatively short passage. It is an economical on one hand, and on the other, it permits the tester to test a number of sub-skills of reading6 through a cloze text.
Transfer of Information
The candidates are required to read a text and complete the information of a chart, form, diagram, table, map, and picture etc. based on what is given in the text. These activities are authentic by nature: people do them in real life.
Unique answer/Short answer
The candidates read a text and answer the questions given. In traditional sense, the questions require candidates to write long answers, which demonstrate not only their reading ability but also writing performances. In ‘Unique Answer’ technique, only one possible correct response is available and the answer is usually a single word, a number etc.
The ‘Unique Answer’ technique is useful because the questions are objective and requires minimal demand on writing. However, its use is limited since it may not allow tester to set many questions for which only one answer is possible in a short text. Therefore, testers prefer to use ‘Short Answer’ technique where the candidates are required to write little longer answers than a single word answer. Even with short answers, the candidates are likely to write ‘correct’ answers with grammatical errors. If marks are deducted for errors of grammar, the sole purpose of testing reading is lost.
One of the commonest weaknesses of the technique is that the examiner cannot judge accurately whether the candidate has arrived at the correct answer. Therefore, the tester can alter this technique by eliciting the desired response through framing the item where the candidate has to complete a sentence as follows;
According to the text, many people in Nelumwila travel to their workplaces by …………
Ordering
The candidates are required to re-arrange the given sentences to form a meaningful text as in the following question:
In spite of this damage the two ships managed to reach the nearby port of Durnkirk under their own steam.
However, because of the weather conditions, the captains did not realize the danger until only second before the collision took place.
One was a cargo boat carrying fruit, and apparently bound for London.
Fortunately, there were no casualties among the crews or the passengers but both ships were holed closed to the water-line.
According to eyewitness neither of the vessels was going very fast at the time of the accident.
Yesterday evening two ships collided in thick fog in the English Channel.
The other was a British Rail cross channel ferry on its regular run from Dover to Dunkirk.
The test item tests candidate’s ability to relate one part of the text to another, the linkers etc. The defect in this type of test item is that if a candidate puts one item in a wrong position in the order, it can affect the others as well.
Do'stlaringiz bilan baham: |
