ENG515 Quiz Lectures 23-45

Lesson-23 · ASSESSING READING II · Topic-127: Reading Aloud

Reading aloud can be used to check the learner’s skill at word recognition. As a very cautious first step it is worth observing carefully to see if the learner’s eyesight is good. This could be done by getting them to look at a picture and then asking them questions about it. Quite a large proportion of males are color-blind to some degree, but that should not affect reading. If the learner seems to have eyesight problems, it is worth getting their eyesight tested by a specialist.

Reading aloud should begin with a very easy short text. If the learner has problems in reading aloud very early in the text, it may be worthwhile pausing and talking about the context of the story with the learner, discussing some of the ideas that will occur in the text and predicting what might happen in the story. It is probably not worth keeping a running record of errors for the first text, but if it becomes clear that word recognition is a major problem, then keeping such a record would be useful. If the learner has some problems with word recognition these could be checked against the correspondences in Appendix 1 to see if they are irregular items or if there is some pattern to the errors.

A difficulty with reading aloud for second language learners is that their skill in reading may be greater than their skill in speaking and so their spoken production may be a poor representation of their reading. Talking to the learner before the reading begins is one way of checking this.

Topic-128: Vocabulary Tests

Learners may have difficulty reading because they do not know enough vocabulary. Note that word recognition during reading aloud is affected by vocabulary knowledge and so very easy texts need to be used at first when testing reading aloud. Similarly, if the vocabulary test is a written test which requires the learners to read the test items, then the measure of vocabulary knowledge will be affected by word recognition skills. Learners may have a large spoken vocabulary but be unable to read the words they know. The following vocabulary tests can be used with learners of English.

The Bilingual Levels Tests

Here is an item from the Indonesian version of the test (available in Nation, 2004b).

1 could

2 during dapat, bisa

3 this selama

4 piece supaya

5 of

6 in order to These are tests of the first and second 1,000 words of West’s (1953) General Service List. They are available in the following languages—Japanese, Indonesian, Thai, Korean, Chinese (traditional and simplified), Tagalog, Samoan, Tongan, Russian, and Vietnamese. The test is easily marked with a marking key. Each of the two levels has 30 items, which is enough for a good level of reliability. The validity of the test is strengthened for low proficiency learners by the use of the first language to represent the meanings of the words. The learners do not have to deal with the more complex language of English definitions. The words are tested out of context and this can cause problems for words that can have different meanings, such as seal which can mean “to close tightly” or “the marine mammal”, or bear which can mean “to carry” or “to put up with”. The teacher should sit next to the learner while the test is being done to make sure that the learner takes the test seriously, follows a sensible test-taking strategy, knows how to handle the slightly unusual test format, and is not experiencing reading problems which might interfere with the sitting of the test. It is very important that the teacher does this because it has happened that a whole class of learners sitting a test has not taken it seriously and has got low marks. In this case, the teacher then set up a programme to teach the vocabulary which the learners actually already knew.

The True/False Vocabulary Test

Here are three sample items (available in Nation, 2001: 412–415; Nation, 2004b; Nation, 1993).

When something falls, it goes up.

Most children go to school at night.

It is easy for children to remain still.

There are two versions of this 40-item test of the first 1,000 words of West’s (1953) General Service List. The test can be given in a written form or, if necessary, it can be given orally. There are enough items to get a good level of reliability. The words are tested in sentences and learners need to understand the sentence and apply it to their knowledge of the world in order to make a decision about whether the sentence is true or false. As there are factors other than vocabulary knowledge involved in the test, this affects the validity of the test. Whether this effect is positive or negative depends on what you want to test, words alone or words in use. Other possible tests include the yes/no test, and the monolingual levels test (Schmitt, Schmitt and Clapham, 2001).

Topic-129: Tests of Grammatical Knowledge

If the teacher speaks the first language of the learners, the most straightforward test of grammatical knowledge is to get the learners to translate sentences from reading texts, starting with a very simple text. A validity issue with this is that such translation may encourage word by word reading and as a result mistranslation. This can be discouraged by asking the learner to read the whole sentence first before beginning the translation. An example of mistranslation is “He made the theory useful” being translated as “He made the theory which was useful”. If the teacher does not speak the learners’ first language, sentence completion tests could be used, for example,

I was very surprised by _________________.

It made me___________.

I ________________waiting for at least an hour. Note that grammar tests, both translation and completion, involve word recognition skills and vocabulary knowledge as well as grammatical knowledge. It is thus important that the learners’ word recognition skills and vocabulary knowledge are tested before grammar knowledge is tested.

Topic-130: Measuring Reading Proficiency

A proficiency test tries to measure a learner’s skill not in relation to what has been taught on a particular course but in relation to a wider standard. Tests like the TOEFL test and IELTS test are proficiency measures. New Zealand has reading proficiency tests for school-age native speakers of English in the Progressive Achievement Tests (PAT) series.

Typically such tests use multiple-choice questions with several texts. In some tests, cloze tests may be used but these are not so popular, probably for reasons of face validity rather than because of the effectiveness of the tests.

We have looked at multiple-choice tests in the section on achievement tests. We will look at cloze tests first in detail in this section.

Cloze Tests

Here is an example of a cloze test. The complete text is given in Figure 6.1 later in this chapter.

Return Ticket, Please!

by David Hill

The first person ever to stand on the moon was Neil Armstrong on 20 July 1969.

As he stepped out (1) his spacecraft, he said, “(2) ’s one small step (3) a man; one giant (4) for mankind.” What would (5) say if you were (6) first person ever to (7) onto another planet? The (8) States, working with other (9) , is planning a new (10) to Mars by the (11) 2020. And this time, (12) ’ll be people on (13) . Spacecraft without crews have (14) landing on Mars since (15) 1970s. In 1997, the (16) Pathfinder landed on Mars, (17) a small, sixwheeled (18) called Sojourner trundled around, (19) rocks and measuring the (20) speed. A trip to (21) takes a long time. (22) astronauts who walked on (23) moon took three days (24) get there, travelling at (25) km/h in their Apollo (26) . At the same speed, (27) would take them twenty (28) to reach Mars. Even (29) today’s more powerful rockets, (30) could take at least seven months to reach Mars. (adapted from The School Journal 2000: Part 2)

Every fifth word has been taken out of a reading passage that the learners have never seen before. The learners must fill in the missing words by guessing. They look at the words before and after the empty space to help them guess the missing words. The test measures how close the reader’s thought is to the writer’s thought.

Usually in a test like this, there need to be 40–50 empty spaces to reach a good level of reliability. The words must be taken out according to a plan. Every fifth word can be left out, or every sixth or seventh, etc., but it must be done in a regular way. A line is drawn to show each missing word. Usually the first sentence in the text has no words removed. There are two ways of marking. One way is to accept any sensible answer (acceptable alternative). Another way is to accept only the words that are exactly the same as the ones left out (exact