Summary

Posted: October 17th, 2013

Name:

Course:

Instructor:

Date:

Summary

Kinds of Tests and Testing

Different types of tests exist, and they include proficiency, achievement, diagnostic, and placement tests. Proficiency tests measure a person’s ability in a language, irrespective of their training in that language. The types of proficiency tests differ in the content and level of difficulty, but they are similar in the fact that they do not depend on the students’ qualifications. Proficiency tests can be general, with the aim of showing that the students have acquired a predetermined standard, or they can be taken in a specified area. Achievement tests gauge the students’ performance, after having taken a language course. The students take final achievement tests after the end of the course. They can be in the form of syllabus content approach, where the administrators use the course syllabus, or they can be based on the objectives of the course. Progressive achievement tests help teachers to determine the progress that the students are making in the course of their study.

Diagnostic tests identify the areas that students need to improve, by checking the strengths and weaknesses of the students. It is often difficult for teachers to design diagnostic tests, because of the vast information required. Most of the diagnostic tests contain unreliable information, and some are highly detailed. The implementation of technology, specifically computers with large memories, is an added advantage when administering these tests, as they will lessen the time taken to administer and use diagnostic tests. Teachers use placement tests to determine the level at which they will place the students. The teachers have to know the features at each teaching level. This will help them in designing the most appropriate tests, and will help place the student in the most appropriate level. Teachers can use direct or indirect testing when constructing tests. When using direct testing, the teachers identify areas where they want to examine the students and test them. When using indirect testing, teachers identify a skill, and then they test the students’ abilities in acquiring that skill. Teachers can use discrete point testing, where they test the students using one element at a time, or they can use integrative testing, where they test the students using different elements.

Questions

1. Identify and discuss the different forms of tests. How do these tests help in second language evaluation, and which test, in your opinion, is the most appropriate in gauging a student’s performance?
2. Technology is important and useful in every sector. How can administrators use technology in second language evaluation? Of what purpose is technology when using diagnostic tests?

Validity

Validity ensures the relevance of the tests. There are different forms of validity, and they include content validity and criterion-related validity. When checking whether tests have content validity, teachers compare the tests specifications and the test content. Content validity is determined during test construction, to ensure that the tests contain all areas of specification. Content validity enhances construct validity. Tests with high content validity help in ensuring more accuracy. High content validity ensures that the major areas in testing are covered fully. In addition, high content validity ensures the reduction of harmful backwash effect. It ensures that teachers do not ignore any areas when teaching. Criterion-related validity ensures that the results of a test relate with the result of another related but independent and reliable test. The teachers use the independent test as a criterion for the test. Criteria-related validity can be concurrent or predictive. Teachers should ensure that the tests and the scores are valid.

Face validity determines validity based on appearances. It is possible for tests to lack face validity even when they have content and criteria-related validity. Face validity is crucial, even though it does not enhance construct validity. Tests lacking face validity are often rejected by all the participants. Students who take such tests, often perform below their abilities. It is essential to perform a validation exercise before issuing a test. Teachers find it challenging to ensure full validation when they construct the tests. Teachers and other test developers can undertake several measures to ensure that the tests they construct are valid. The first thing is to ensure that the tests contain all the information and specifications on what the teachers will measure. They should use direct testing, and in case they have to use indirect testing, they should provide references showing that the tests have been validated. They should ensure that the scores are in direct relation to the test. They should ensure that their tests are reliable since a tests lacking reliability lacks validity.

Questions

1. How do content validity and criterion-related validity differ, and how do they help in enhancing construct validity
2. How is it possible for teachers to determine that tests have content validity and criterion related validity, when the same tests lack face validity? How does lack of face validity affect the students?

Reliability

Reliability refers to the consistency and stability of the tests. Reliable tests produce almost similar scores when the same test is undertaken under different times by the same students, having the same abilities. Reliability coefficients enable teachers to quantify test reliability, using different tests. There are different ways of obtaining reliability coefficients. The test-retest method involves the students taking the same exam twice. Reliability is enhanced when the students do not stay for long periods in between the exams. The split-half method requires students to take the exam once, but the teacher assigns two scores. The first score represents the first half of the test, and the second score is for the second part. The ideal reliability coefficient is one. A test with this coefficient means that the test would give the exact results regardless of the time when the test was administered. A coefficient of zero is not desired, as the results would differ greatly when the students take it at different times.

Different types of language tests have identifiable reliability coefficients. Reliability coefficients for good vocabulary, structure, and reading are high, and they range between 90 and .99. Coefficients for auditory comprehension tests range between .80 and .89. Coefficients for oral production tests are lower than the other two tests, and they range between .70 and .79. The above scores are a reflection of the difficulties encountered when attaining for reliability in tests. Reliability will depend on the importance of the decision to be undertaken after the test. If the decision to be undertaken is of great importance, then the tests must be of considerable reliability. Reliability coefficients determine the standard error of measurement. The standard error of measurement is small when there is a great reliability coefficient. Test reliability is enhanced when there are many independent items on a test. Strong and weak students should perform differently on similar items, as this ensures reliability.

Questions

1. What is the purpose and importance of the standard error of measurement, and how does it differ from the reliability coefficient?
2. What is the importance of constructing a longer test, with more independent items? How do such tests help the students?

Name:

Course:

Instructor:

Date:

Summary

Kinds of Tests and Testing

Different types of tests exist, and they include proficiency, achievement, diagnostic, and placement tests. Proficiency tests measure a person’s ability in a language, irrespective of their training in that language. The types of proficiency tests differ in the content and level of difficulty, but they are similar in the fact that they do not depend on the students’ qualifications. Proficiency tests can be general, with the aim of showing that the students have acquired a predetermined standard, or they can be taken in a specified area. Achievement tests gauge the students’ performance, after having taken a language course. The students take final achievement tests after the end of the course. They can be in the form of syllabus content approach, where the administrators use the course syllabus, or they can be based on the objectives of the course. Progressive achievement tests help teachers to determine the progress that the students are making in the course of their study.

Diagnostic tests identify the areas that students need to improve, by checking the strengths and weaknesses of the students. It is often difficult for teachers to design diagnostic tests, because of the vast information required. Most of the diagnostic tests contain unreliable information, and some are highly detailed. The implementation of technology, specifically computers with large memories, is an added advantage when administering these tests, as they will lessen the time taken to administer and use diagnostic tests. Teachers use placement tests to determine the level at which they will place the students. The teachers have to know the features at each teaching level. This will help them in designing the most appropriate tests, and will help place the student in the most appropriate level. Teachers can use direct or indirect testing when constructing tests. When using direct testing, the teachers identify areas where they want to examine the students and test them. When using indirect testing, teachers identify a skill, and then they test the students’ abilities in acquiring that skill. Teachers can use discrete point testing, where they test the students using one element at a time, or they can use integrative testing, where they test the students using different elements.

Questions

3. Identify and discuss the different forms of tests. How do these tests help in second language evaluation, and which test, in your opinion, is the most appropriate in gauging a student’s performance?
4. Technology is important and useful in every sector. How can administrators use technology in second language evaluation? Of what purpose is technology when using diagnostic tests?

Validity

Validity ensures the relevance of the tests. There are different forms of validity, and they include content validity and criterion-related validity. When checking whether tests have content validity, teachers compare the tests specifications and the test content. Content validity is determined during test construction, to ensure that the tests contain all areas of specification. Content validity enhances construct validity. Tests with high content validity help in ensuring more accuracy. High content validity ensures that the major areas in testing are covered fully. In addition, high content validity ensures the reduction of harmful backwash effect. It ensures that teachers do not ignore any areas when teaching. Criterion-related validity ensures that the results of a test relate with the result of another related but independent and reliable test. The teachers use the independent test as a criterion for the test. Criteria-related validity can be concurrent or predictive. Teachers should ensure that the tests and the scores are valid.

Face validity determines validity based on appearances. It is possible for tests to lack face validity even when they have content and criteria-related validity. Face validity is crucial, even though it does not enhance construct validity. Tests lacking face validity are often rejected by all the participants. Students who take such tests, often perform below their abilities. It is essential to perform a validation exercise before issuing a test. Teachers find it challenging to ensure full validation when they construct the tests. Teachers and other test developers can undertake several measures to ensure that the tests they construct are valid. The first thing is to ensure that the tests contain all the information and specifications on what the teachers will measure. They should use direct testing, and in case they have to use indirect testing, they should provide references showing that the tests have been validated. They should ensure that the scores are in direct relation to the test. They should ensure that their tests are reliable since a tests lacking reliability lacks validity.

Questions

3. How do content validity and criterion-related validity differ, and how do they help in enhancing construct validity
4. How is it possible for teachers to determine that tests have content validity and criterion related validity, when the same tests lack face validity? How does lack of face validity affect the students?

Reliability

Reliability refers to the consistency and stability of the tests. Reliable tests produce almost similar scores when the same test is undertaken under different times by the same students, having the same abilities. Reliability coefficients enable teachers to quantify test reliability, using different tests. There are different ways of obtaining reliability coefficients. The test-retest method involves the students taking the same exam twice. Reliability is enhanced when the students do not stay for long periods in between the exams. The split-half method requires students to take the exam once, but the teacher assigns two scores. The first score represents the first half of the test, and the second score is for the second part. The ideal reliability coefficient is one. A test with this coefficient means that the test would give the exact results regardless of the time when the test was administered. A coefficient of zero is not desired, as the results would differ greatly when the students take it at different times.

Different types of language tests have identifiable reliability coefficients. Reliability coefficients for good vocabulary, structure, and reading are high, and they range between 90 and .99. Coefficients for auditory comprehension tests range between .80 and .89. Coefficients for oral production tests are lower than the other two tests, and they range between .70 and .79. The above scores are a reflection of the difficulties encountered when attaining for reliability in tests. Reliability will depend on the importance of the decision to be undertaken after the test. If the decision to be undertaken is of great importance, then the tests must be of considerable reliability. Reliability coefficients determine the standard error of measurement. The standard error of measurement is small when there is a great reliability coefficient. Test reliability is enhanced when there are many independent items on a test. Strong and weak students should perform differently on similar items, as this ensures reliability.

Questions

3. What is the purpose and importance of the standard error of measurement, and how does it differ from the reliability coefficient?
4. What is the importance of constructing a longer test, with more independent items? How do such tests help the students?