Language testing and methods of assessment

Methods of Assessment

Reasons for Testing: The Objectives of Different Test Types

The questions below introduce some of the main issues involved in language assessment. Consider each question before reading the answer.

• Try to formulate your own answer before looking at the discussion.
• Then click Check your answer to go to the relevant section.

1. Check your answer: What basic questions should a teacher or tester ask before setting a test?
2. Check your answer: What are the relationships between learning, teaching and testing?
3. Check your answer: What are the main reasons for testing?
4. Check your answer: What are the essential differences between a classroom test and an external examination?
5. Check your answer: Briefly describe the following types of test in terms of their objectives: (a) aptitude, (b) placement, (c) diagnostic, (d) achievement and (e) proficiency.
6. Check your answer: How far is it possible to predict student success in language learning?
7. Check your answer: Can a student's progress in learning a second language (L2) be measured? What are the difficulties?
8. Check your answer: How does one specify test objectives? How far does the way objectives are defined influence subsequent assessment?
9. Check your answer: What are the main language skills and their components? Are there other skills that should be tested?

1. What basic questions should a teacher or tester ask before setting a test?

Key questions: What should be tested? How should it be tested? Is testing necessary? Why is the assessment being made? What should the test contain? What will be the consequences for teaching, learning and administration? How good is the proposed test material?

The characteristics of a good test include:

• Validity – it should measure what it is intended to measure and, as far as possible, nothing else.
• Reliability – it should produce reasonably consistent results. If it were administered again, without intervening learning or teaching, a reliable test should produce a broadly similar order of merit.
• Discrimination – the tester should first decide whether the purpose of the test requires candidates to be differentiated from one another. School examinations, for example, are generally designed to discriminate among candidates.
• Backwash or washback – the test inevitably influences teaching and learning. An important question is whether its influence on the learning and teaching which precede the test is beneficial.

2. What are the relationships between learning, teaching and testing?

Content validity involves ensuring that the purposes of assessment reflect the aims of the teacher and learner.

Content specification involves identifying what is to be tested. It should ensure that the test reflects the areas to be assessed in appropriate proportions.

A balanced sample of test items is therefore important. The nature and distribution of the items should reflect the objectives of the assessment.

Backwash or washback refers to the influence of testing on teaching and learning before the test. In this sense, testing can become part of the teaching process: what is tested tends to influence what teachers teach and what learners study.

3. What are the main reasons for testing?

Different types of test serve different purposes.

Achievement or attainment tests: These are usually more formal and are designed to show mastery of a particular syllabus. Examples include end-of-year tests, school-leaving examinations and public examinations. Although they may resemble progress tests, they are normally related to a defined syllabus and are designed primarily to measure achievement rather than to motivate or reinforce learning.

Progress tests: Most classroom tests take this form. They assess the progress students have made in mastering material taught in the classroom. They can motivate students, help teachers and learners assess the success of teaching and learning, and identify areas of weakness or difficulty. Progress tests can therefore also have a diagnostic function.

Diagnostic tests: These are designed to identify particular weaknesses or difficulties so that an appropriate programme of remedial work can be planned. A diagnostic test may contain elements of progress, achievement or proficiency testing, but its defining purpose is to identify what a learner can and cannot do. It may be administered before a course begins or at an appropriate point during a course.

Placement tests: These are used to place new students in teaching groups whose members are approximately at the same level when they begin a course. They are concerned with students' present level of ability rather than with mastery of particular points in a known syllabus. A variety of tests may therefore be required. Placement tests need to produce results quickly because they also serve an important administrative purpose.

Proficiency tests: These measure whether students have the language ability required for a particular future purpose, such as following a university course through English or performing a particular job. They therefore look forward to the application of language rather than back to the syllabus through which the learner has studied. Proficiency tests should define the operational language needs of the task and may involve practical or authentic situations. They apply a common standard regardless of the candidate's previous learning.

Aptitude tests: These attempt to predict probable performance in language learning. They look forward to future performance but differ from proficiency tests. Aptitude tests are concerned with a learner's potential for learning or using language—for example, whether the learner is likely to experience difficulty identifying the sounds or grammatical structures of a new language. Proficiency tests, by contrast, measure the adequacy of a learner's existing control of L2 for carrying out a particular non-language task.

4. What are the essential differences between a classroom test and an external examination?

One important difference concerns purpose. Most external examinations are designed to discriminate among candidates, often so that candidates can be ranked, graded or selected.

A classroom test, by contrast, may have a more immediate teaching and learning purpose. It may be used to monitor progress, diagnose difficulties, provide feedback or motivate students. The same test format can therefore have very different implications depending on why the assessment is being made.

5. What are the main types of test in terms of their objectives?

• Aptitude: looks forward and concerns probable future language performance, rather than achievement in a known course.
• Placement: places new students in teaching groups at an appropriate level.
• Diagnostic: identifies areas of weakness or difficulty, usually with a view to remedial teaching.
• Achievement: relates to a known syllabus and assesses the extent to which its objectives have been achieved.
• Proficiency: looks forward to the use of language for a particular non-language purpose, such as university study or employment.

6. How far is it possible to predict student success in language learning?

Aptitude tests attempt to assess proficiency in language for language use—that is, the learner's potential for language performance itself. They do not assess the use of language to undertake a particular non-linguistic task.

There are several problems in attempting to predict success.

First, what do we mean by success in language learning? The answer depends partly on the purposes for which the language is being learned.

Second, before we can predict success, we need to be able to measure it. This raises the difficult question of how language proficiency itself should be measured.

Third, linguistic skill is unlikely to be the only factor involved. Even if an aptitude test successfully measures aspects of linguistic ability, other factors may contribute to successful performance.

It is difficult to separate linguistic skill from factors such as intelligence, motivation and the teaching situation. There are therefore many variables involved, and further research is needed into what is fundamentally involved in language learning.

A further question is how far aptitude can be separated from general intelligence. There may be a relationship between high IQ and success in language learning, but this does not necessarily mean that the two are the same thing.

7. Can a student's progress in learning a second language be measured?

The first requirement is a clearly defined syllabus: a clear statement of aims and methods together with a specification of the content to be learned.

The design and content of a progress test should then seek to show whether students have attained the abilities which the course is intended to develop.

There are, however, difficulties in attributing progress to the course itself. Progress may also result from previous knowledge, outside influences, time spent in the host environment, independent study or other learning experiences. The quality of the test and the assumptions on which it is based are also important.

8. How does one specify test objectives?

The best basis for setting valid tests is to ask questions at every stage of the test-development process, especially at the beginning. The specification should provide as clear a statement as possible of why the assessment is being made, what it will contain and what consequences it may have for teaching, learning and administration.

Linguistic competence

One approach is to specify objectives in terms of levels of linguistic competence. These may include:

• Grammar and usage – morphology and syntax;
• Vocabulary – lexis, semantics and collocations;
• Listening comprehension – including phonological discrimination and recognition;
• Pronunciation – including sounds, stress and intonation;
• Writing – including graphology and written language skills.

9. What are the main language skills and their components?

The four traditional language skills are listening, speaking, reading and writing. Each can, however, be divided into a range of more specific skills and sub-skills. In addition, particular language tasks may require learners to draw on a combination of skills.

For detailed lists of skills and sub-skills, see:

• Communicative Syllabus Design, John Munby (Cambridge University Press, 1978), especially the Taxonomy of Language Skills, pp. 123–131. Examples include discriminating and articulating sounds in isolated words and connected speech; discriminating and articulating stress patterns; producing intonation patterns; recognising the script of a language; deducing the meaning of unfamiliar items; and understanding conceptual meaning and communicative value.
• Developing Reading Skills, Françoise Grellet (Cambridge University Press, 1981). This is a particularly accessible introduction to the teaching of reading.
• Comprehension and Comprehension Tests, Lunzer et al. (1979).
• Research in Comprehension in Reading, pp. 499–545, F. B. Davis (1968).
• Reading Research Quarterly, Vol. 3: "Identification of Subskills in Reading Comprehension by Maximum Likelihood Analysis", D. Spearritt (1972).
• Reading Research Quarterly, Vol. 8: "A Model Program for Teaching Advanced Reading to Students of EFL", D. E. Eskey.
• Language Learning, Vol. 23, No. 2, pp. 169–184: "Reading as Problem Solving: An Investigation of Strategies", J. E. Olshavsky.
• Industrial English, T. Jupp and S. Hodlin (1975), dealing with the theory and practice of functional language teaching.
• B. J. Carroll's Operational Specifications, against which candidates' performance can be matched:
  1. size and complexity of the text;
  2. range of enabling skills, structures and functions handled;
  3. speed at which language can be processed;
  4. flexibility in dealing with changes of topic;
  5. accuracy of structures;
  6. appropriacy of functions;
  7. independence from reference sources;
  8. repetition and hesitation in processing the text.
• ESU Framework, Carroll, B. J. & R. West (Longman, 1989). This contains operational objectives for language skills and sub-skills and attempts to chart the standards set by different British examination boards within a single framework. It includes specifications for testing all four skills and more.