Tập huấn Thay sách Anh 12 - 2008

TESTING

Presented by
TESTING & ASSESSMENT
Questions for discussion
1.State what you know about innovation/change in testing and assessment.
2.List some test types you usually give your 12-grade students.
3.Suggest what should be done to make fair tests.
Follow-up activity
In groups, choose one point to be tested (either language skills or vocabulary or grammar or pronunciation). Design a 45-minute test with key answers.

Innovation/Change in Testing
In traditional language teaching, formal and informal tests are given to students by their teachers.
Nowadays, based on student-centered approach, language testing includes tests, continuous/continual assessments.
- Formal tests are given to students at the end of a semester/ school year, marked, and assessed by teachers.
- Continuous/continual assessments are informal tests given regularly to students during the course. They are assessed by teachers( teacher assessment), or classmates (peer assessment), or the student himself/ herself (self-assessment or individual assessment).




CHARACTERISTICS OF A GOOD TEST

VALIDITY & RELIABILITY
Validity
A test is valid
- if it tests what is supposed to test (e.g. test writing ability with essay question that requires specialist knowledge of history or biology)
- if it produces similar results to some other measure (i.e. Test A gives the same kind of resuts as Test B).
- when there is validity in the way it is marked(e.g. test listening ability but scoring the answers for spelling and grammar  not a valid test of listening)






VALIDITY & RELIABILITY
Reliability
A good test should give consistent results (i.e. same test same results on each occasion).
Reliability depends on
making the test instructions clear;
restricting the scope for variety in the answers;
ensuring test conditions remain constant;
the scorers in particular (i.e. a test is unreliable if the result depends to any large extent on who is marking it.)


Testing and Assessment
“…Testing is one kind of assessment. One which is typically used at the end of a stage of instruction to measure student achievement. Assessment is a broader concept: it is part of the whole educational process of teaching and learning…”
“…Whilst tests can be used as a ‘bolt-on’ procedure at end-points in a learning program, assessment is integral to the whole process of teaching and learning…”
Assessment is undertaken and for different purposes. The first purpose is pedagogically motivated, i.e. formative assessment. The second one is to measure learner achievement, i.e. summative assessment.
Source: Hedge, T. (2000). Teaching and Learning in the Language Classroom, OUP .(pp 336-337)
Formative versus Summative Assessment
Source: Hedge, T. (2000). Teaching and Learning in the Language Classroom, OUP .(pp 336-337)


Formative assessment
is prepared and carried out by the class teacher as a routine part of teaching and learning.
is specifically related to what has been taught, i.e. content is in harmony with what has been taught.
the information from the assessment is used diagnostically; it focuses on the individual learner’s specific strengths and weaknesses, needs, etc.
Summative assessment
is not necessarily prepared and carried out by the class teacher.
does not necessarily relate immediately to what has been taught.
the judgement about a learner’s performance is likely to feed into record-keeping and be used for administrative purposes, e.g. checking standards and targets.
is frequently externally imposed, e.g. by institution or a ministry of education.

Some test types

Teachers might use a variety of test types which will depend on
The level of students
The skills they want to test
The language points they want to test
A. Listening tests
1. Multiple choice questions
2. True-False statements
3. Open-ended questions
4. Grid-filling/ Completing charts with facts & figures
5. Gap-filling
6. Identifying objects/people/speakers’ tones

Some test types (cont.)
B. Reading tests
1. Multiple choice questions
2. True-False statements
3. Open-ended questions
4. Grid-filling
5. Gap-filling/ Cloze
6. Matching (jumbled headings with paragraphs/written descriptions with pictures of the items or procedure they describe)
7. Transferring written information to charts, graphs, maps,etc.
8. Choosing the best summary of a paragraph or a whole text.


Some test types (cont.)
C. Speaking tests
1. Prompts/Picture prompts
2. Asking & answering questions(Interview)
3. A set of topics
4. Role-plays
5. Information gap activities (e.g. describe and draw/ describe and arrange)
6. Decision-making activities (e.g. show ten photos of people and ask candidates to put them in order of the best and worst dressed.)
7. Using pictures for candidates to compare and contrast (e.g. to find similarities & differences)


Some test types (cont.)
D. Writing tests
Guided writing tests:
- Gap-filling
- Punctuating a text
- Re-ordering words/sentences
- Summary writing based on texts
Free writing tests:
- Sentence writing (e.g. a set of instructions for some common task)
- Paragraph writing (e.g. writing leaflets/brochures)
- Letter writing (e.g. transactional letters)
- Essay writing (e.g. writing compositions and stories)



Types of test item
A test item is direct if it asks testees to perform the communicative skill which is being tested. Direct test items try to be as much like real-life language as possible.
Indirect test items try to measure a student’s knowledge and ability through more controlled items such as multiple-choice questions or grammar transformation items.

Some suggestions for making fair tests
Before designing a test and giving it to students teachers should take these things into considerations:
Assess the test situation.
Teachers should remind themselves of the context, the allotted time, the place test will take place, time for marking.
Decide what to test.
Language skills/vocabulary items/ grammar points/ appropriate topics?





Some suggestions for making fair tests
Balance the elements.
Balancing elements involves estimating how long teachers want each section of the test to take and then writing test items within those time constraints.
Weigh the scores.
Give more marks to more important questions.
Make the test work.
Try out individual items or whole tests on colleagues or other students before administering them to the real testees.
Some suggestions for making fair tests

Ask enough questions to test accurately.
A test will be less reliable if it has too few questions. With more questions, one question will not cause too much a problem.
Write clear instructions/directions for the test and use examples.
Clear instructions will help students understand the questions, know how to answer, and answer the questions accurately.



Some suggestions…
Test what has been taught in class.
A test based on what has not been taught in class would not be a reliable test.
Write clear test items.
Compare the following:
Item 1: Directions: Choose the best response.
Today it is rather cloudy.
A. It will probably rain.
B. Yesterday the weather was nice.
C. Yes, I think so, too.

Some suggestions…
Item 2: Directions: Choose the letter for the answer which means the same as the statement.
Today it is rather cloudy.
A. The clouds are dark.
B. It is not too sunny today.
C. It might snow tomorrow.

Which item is clear?
Item 2 is clear; Item 1 is unclear.
Some suggestions…
When writing multiple choice questions, use realistic distractors.
Distractors are wrong answers. A test will not be reliable if impossible distractors are used.
Consider the following question:
What color is the sun?
A. Yellow ( This is intended correct answer)
B. Orange ( too close to the correct answer)
C. Soft (not a color & too far from the correct answer)
D. Yes (It is a bad distractor because this is an answer to a yes/no question)
A better set of answers to test knowledge of colors would simply be a list of colors:
A. blue B. green C. yellow D. black
Some suggestions…
Score tests consistently with a key.
When scoring, use a key that will help teachers grade objectively. The key should have examples of correct answers ( though other correct answers should be accepted) and the appropriate points/marks as well.

Avoid traps for students
A good test should never be constructed in such a way as to trap the students (testees) into giving an incorrect answer. All tests should be constructed primarily with the intention of finding out what a student knows _ not with trapping him.