FULL 038
DEVELOPING SPECIFIC ENGLISH LANGUAGE PROFICIENCY TESTS
(SELProTest) FOR INDONESIAN CONTEXTS
Dr. Y. M. Harsono
ym.harsono@atmajaya.ac.id
Atma Jaya Catholic University Jakarta
INTRODUCTION
Observing English language tests used all over the world, there are at least five types of test: progress test, achievement test, diagnostic test, predictive test, placement test, and proficiency test each of which has its own function. Progress test, for example, functions to measure the progress made by students towards defined goals the contents of which are generally related to the detailed course objectives. Proficiency test is intended to measure how much of a language someone has learned without necessarily referring to a particular course of instruction. TOEFL and IELTS are two most well-known English proficiency tests for those who intend to study to the universities in the US and England or Australia respectively.
The problem is whether students who intend to learn and people who want to work in Indonesia have to take either TOEFL or IELTS. These tests are probably suitable to identify the general English proficiency for the test takers and not suitable for the specific one. More than that, taking these tests is costly and cannot be done anytime needed.
The objective of this paper is to share the idea of developing SELProTest which is more suitable to identify the test takers specific English proficiency. This paper, therefore, will discuss communicative language test (CL Test) and English language proficiency test.
COMMUNICATIVE LANGUAGE TEST
This first major section of CL Test elaborates its concept, characteristics, construction, and methods.
The Concept of CL Test
The concept of CL Test argues that it should not only test the learners’ competence, that is, what the learners know about the second language and about how to use it but also to the performance, that is, to what extent the learners are able to actually demonstrate this knowledge in a meaningful communicative situation.
The concept of CC would be elaborated through the historical sketch of CC from Dell Hymes (1972), Canale and Swain (1980), Canale (1983), and Celce Murcia et at. (1995). The term “Communicative Competence” was first coined by Dell Hymes (1972) as a reaction against the concept of language competence which was proposed by Chomsky (1965). According to Chomsky competence simply means ‘knowledge of the language system’ or the grammatical knowledge. Chomsky’s view of language competence was more psycholinguistic in nature, whereas Hymes’ concept was more sociolinguistic. Hymes’ concept of competence includes concepts of appropriateness and acceptability.
Hymes divides communicative competence into four sectors. The first, ‘whether or not something is formally possible’ which is equal to Chomsky’s concept of competence (that is grammatical competence). The second sector deals with feasibility, that is, whether a sentence is grammatically possible. The third sector covers appropriateness to context. The fourth sector refers to ‘accepted usage’. It concerns whether or not something is in fact done (Hymes in Brumfit and Johnson, 1987).
Hymes’ term of communicative competence was interpreted and developed in sociolinguistic area by many sociolinguists: Canale and Swain (1980), Canale (1983), and Celce-Murcia et al. (1995). Canale and Swain (1980) state that CC includes grammatical competence, that is, knowledge of the rules of grammar; sociolinguistic competence, that is, the rules of use and rules of discourse; and strategic competence, that is, knowledge of verbal and non-verbal communication strategies. According to Canale (1983) “communicative competence consists of four domains of knowledge and skills, i.e. grammatical, sociolinguistic, discourse and strategic competence”.
Grammatical competence refers to Chomsky’s linguistic competence. It deals with the mastery of verbal and non-verbal linguistic codes. This competence includes the mastery of vocabulary, words, phrases, sentences, and pronunciation. This competence is very much needed to interpret and convey the literal meaning of utterances.
Sociolinguistic competence relates to the ability to select and use suitable linguistic forms for a certain context of communication. This competence also has to do with socio-cultural and discourse rules. Therefore, this sociolinguistic context will depend much on certain factors like the objective of the interaction, the status of the speaker and the listener, and the norms and rules of the interaction.
Discourse competence is related to the mastery or ability of the learners to combine grammatical forms and meaning to form a complete spoken or written discourse or text. The unity of a text or discourse is realized through the use of form cohesion and meaning coherence. Cohesion is the relationship between utterances and grammatical structure devices to help one to interpret the meaning of a discourse. Coherence is the relationship among several meanings in a text.
Strategic competence is the ability to use verbal and non-verbal communication strategies in order to make-up for the weaknesses in communication because of the limitation of the language mastery. In addition to that this strategic competence is also used to strengthen the effectiveness of communication.
Celce-Murcia et al. (1995) proposed more detailed model of CC consisting of five components: (1) discourse competence, (2) linguistic competence, (3) actional competence, (4) sociocultural competence, and (5) strategic competence. They divide sociocultural competence of Canale and Swain’s model into three: sociocultural, actional, and discourse competence. The terminological change of grammatical competence to become linguistic competence is to indicate that this competence also includes lexis and phonology in addition to morphology and syntax. Actional competence is defined as “competence in conveying and understanding communicative intent, that is, matching actional intent with linguistic form based on the knowledge of an inventory of verbal schemata that carry illocutionary force (speech acts and speech acts sets)” (Celce-Murcia, 1995:17).
The Characterisics of CL Test
The characteristics of CL Test include focusing on context, authenticity, purpose, direct and integrated test, and holistic and qualitative assessment (Weir, 1990).
To discuss the characteristics of CL Test, one needs to refer back to the communicative approach (CA) of teaching a language. In the implementation of the CA of teaching a language communication is the most important concern of the teaching. The objective of the teaching then is the mastery of the communicative competence (CC). The mastery of CC requires the understanding of the kinds of competence that make up the CC: grammatical, sociolinguistic, discourse, and strategic competence. The complete mastery of the CC includes the mastery of the use, the communicative functions, of language in real communication to serve various communicative needs. It is this language use that should be the objective of the instruction and as the consequence it is also this language use that has to be tested.
Referring to the language use in real life communication implies that there is a variety of communicative language needs. These needs have to be considered in planning and implementing the teaching-learning activities and therefore in testing as well. In order to cover the variety of communicative needs that have been done in the teaching-learning activities, there are also a variety of testing strategies and formats.
According to Carroll (1980:34-35) there are three test format categories: open-ended, closed-ended, and restricted-response. In open-ended tests, the testee is free to give response in accordance to the communicative task he is required to perform. In closed-ended tests, the testee has no chance of expressing his idea freely in relation to responding the question. What he can do is just selecting what has been provided in the options by the test constructor. Restricted-response tests provide the testee with the opportunity to compose his response, even though he is not as free as that in open-ended tests.
CL Tests can include tests of all the four language skills of listening, speaking, reading, and writing and using a variety of formats (Weir, 1990:42-85). The formats of tests of reading comprehension include multiple-choice questions, short answer questions, cloze procedure, selective deletion gap filling, C-Tests, Cloze elide, and information transfer. Testing listening comprehension formats include multiple-choice questions, short answer questions, and information transfer techniques for testing extensive listening skills; dictation, and listening recall for the testing of intensive listening. The formats for testing writing include editing task for indirect methods for assessing linguistic competence; essay tests, controlled writing tasks and summarizing for the direct testing of writing. The formats for testing speaking include verbal essay, oral presentation, the free interview, the controlled interview, information transfer: description of a picture sequence, information transfer: questions on a single picture, interaction tasks, and role play.
In addition to that as a method of testing that is identified more as direct rather than indirect, Djiwandono (2000) states that CL Test is characterized by some other features that need to be considered in its implementation. These features characterize CL Tests as tests which are developed (1) to create events of interaction between participants, (2) to produce unpredictable forms and content of language, (3) to purposively elicit certain communicative function, and authentic tasks on the part of those taking the test, (4) to represent the domain of the test takers in the texts used in the test, (5) to encourage integration of the language skills in testing, and (6) to give priority to qualitative rather than quantitative assessment, especially in assessing productive abilities, where rating scales may be employed.
CL Test emphasizes not on linguistic accuracy but on the ability to function effectively through language in particular contexts of situation. In relation to this statement, Weir (1990) states that there are six features of CL Test. First, context has a very important role and is advocated in the assessment of communicative language ability. Second, the authenticity of tasks and the genuineness of texts in tests is something that should be pursued though it is problematic in its definition and realization. Third, different tests need to be constructed to match different purposes. Fourth, tests of communicative language ability should be as direct as possible. Fifth, integrated testing of communicative language ability is suggested rather than separate skills testing. Finally, CL Tests need to take the holistic and qualitative assessment of productive skills.
Authentic assessment, according to O’Malley and Pierce (1996:4) is described as “the multiple forms of assessment that reflect student learning, achievement, motivation and attitudes on instructionally-relevant classroom activities.” They give three examples of authentic assessment: performance assessment, portfolios, and student self-assessment.
Performance assessment consists of any form of assessment in which the student constructs a response orally or in writing. The student response in performance assessment can be elicited by the teacher in formal or informal assessment contexts or probably observed during classroom instructional or non-instructional settings. Performance assessment requires students to do complex and important tasks using prior knowledge, recent learning, and relevant skills to solve realistic or authentic problems.
The followings are six characteristics of performance assessment which are proposed by O’Malley and Pierce (1996:5): constructed response, higher-order thinking, authenticity, integrative, process and product, and depth versus breadth. (1) Constructed Response: students construct a response, provide an expanded response, engage in a performance, or create a product. (2) Higher-order Thinking: the student typically uses higher levels of thinking in constructing responses to open-ended questions. (3) Authenticity: tasks are meaningful, challenging, and engaging activities that mirror good instruction or other real-world contexts where the student is expected to perform. (4) Integrative: the tasks call for integration of language skills and, in some cases, for integration of knowledge and skills across content areas. (5) Process and Product: procedures and strategies for deriving the correct response or for exploring multiple solutions to complex tasks are often assessed as well (as or sometimes instead of) the product or the “correct” answer. (6) Depth versus Breadth: performance assessments provide information in depth about a student’s skills or mastery as contrasted with the breadth of coverage more typical of multiple-choice tests.
Portfolio assessment is “a systematic collection of student work that is analyzed to show progress over time with regard to instructional objective” (O’Malley and Pierce, 1996:5). Examples of portfolio entries are writing samples, reading logs, drawings, audio or videotapes, and/or teacher and student comments on progress made by the student. The main features of portfolio assessment are the involvement of students in selecting samples of their own work to show growth or learning over time.
Student self-assessment is a key element in authentic assessment and in self-regulated learning (O’Malley and Pierce, 1996). In self-assessment, students have their own strong motivation and strategy to accomplish specific purposes. Thus, students make their own choices of learning activities, managing the time available for them, making use of the available resources within or outside the classroom. Self-regulated learners also collaborate with other students in exchanging ideas, eliciting assistance when needed, and giving support to their peers. Self-regulated learners monitor their own performance and evaluate their progress and accomplishment.
Further O’Malley and Pierce (1996:12) give eight types of authentic assessments: oral interviews, story or text retelling, writing samples, projects/ exhibitions, experiments/ demonstrations, constructed-response items, teacher observations, and portfolios.
Those eight types of authentic assessment are described as follows. (1) For oral interviews, teacher asks student questions about personal background, activities, readings, and interests. (2) For story or text retelling, students retell main ideas or selected details of text experienced through listening or reading. (3) For writing samples, students generate narrative, expository, persuasive, or reference paper. (4) For projects/ exhibitions, students complete project in content area, working individually or in pairs. (5) For experiments/ demonstrations, students complete experiment or demonstrate use of materials. (6) For constructed-response items, students respond in writing to open-ended questions. (7) For teacher observations, teacher observes student attention, response to instructional materials, or interactions with other students. (8) For portfolio, as described before it is a systematic collection of student work to show progress over time.
CL Tests, like all other test designs, need to possess requirements of a good test especially validity and reliability. Both requirements have to be fulfilled in the application of CL Tests. At least one of the existing validity–content, construct, criterion, or face validity–has to be fulfilled in the application of the test. The use of any one of the existing formula to establish reliability of the test has to be followed.
The Construction of CL Test
The construction of CL Test covers five steps: specifying the objectives and content, specifying the types and forms, writing the test items, revising the test items, and finalizing the test.
Carroll (1980) states that there are three major phases in the construction of a testing system: Design, Development, and Operation. The design phase consists of three sub-phases: description of participants, analysis of communicative needs, and specification of test content. The development phase consists of three other sub-phases: realization of test, trial application, and validation and test analysis. The operation phase also consists of three sub-phases: full-scale application, operational use, and revision of test system.
Carroll and Hall (1985) states that there are four phases in language test construction: design, development, operation, and monitoring. This test construction is different from the previous one in the last component, monitoring. The design section consists of the description of testee(s), specification of settings, needs, and statement of test tasks and topics. The development section covers construction of the draft test, trials of test, and analysis of trials and test revision. The operation section is divided into two subsection, that is, introduction of test for practical use, and making decisions on test information. The monitoring section is elaborated into survey of test administration, establishment of test measurement characteristics, and preparation of test revision schedule
Weir (1990:36-41) states that there are four stages in test construction and validation, namely: design, development, operation, and monitoring. These four stages are also in line with those proposed by Carroll and Hall (1985) which can be summarized in the following paragraphs.
Test Design. In test design, the first thing a test developer has to do is defining carefully the testee(s) for whom the test is being prepared. After defining the testees, he has to find their purpose of using the language (their communicative needs), the settings in which they will use it, and all the activities they will need to carry out. The final product of this phase is a detailed table of specification that will function as a blue print for the subsequent stages of the test construction.
Test Development. At this stage, on the basis of the table of specification produced at the first stage, the drafts of the actual tests are developed. The next step is trying out the draft test on the suitable samples. After being tried out, the test results have to be analyzed to get the indication of its validity, reliability, and its practicality. Item analysis also needs to be carried out to make sure that items and tasks work satisfactorily. It is also valuable to get the opinion from the testees, the lay advisers, and the professionals in relation to the content of the test (appropriateness, level, etc.). The test needs to be revised based on both the qualitative as well as quantitative data which are concerned with its items, validity, reliability, and efficiency.
Operation. This stage deals with the first actual operation of the new test. It involves the preparation, printing and distribution of appropriate information manuals for candidates, receiving institutions, invigilators and markers. The information includes familiarising with the test formats and how the answers have to be recorded by the testees.
Monitoring. Monitoring the results of each test administration should be carefully undertaken and the data of the results be stored for future revisions of the test.
Djiwandono (1996:106-116) states that there are five steps to follow: specifying the objectives and content, specifying the types and forms, writing the test items, revising the test items, and finalizing the test. These steps and the description of each step can be put in detail in the following paragraphs.
Step 1: Specifying the objectives and content. Achievement test should be constructed based on the content and the objectives of the teaching. On the basis of the objectives and content of the teaching, a table of specification is developed. Based on the table of specification a picture of what topics to include in the test can be clearly seen.
Step 2: Determining the types and forms. There are a lot of factors to be considered in determining the types and forms of test: number of participants/testees, the materials to cover, the time to administer the test, the teacher’s ability to develop the test, the practicality of administering the test, the scoring and grading, etc. Essay test is easy to construct but needs extra time and concentration to score the test. Objective test, on the other hand, is difficult to construct but very easy to score. The strengths and weaknesses of each test format has to be carefully considered.
Step 3: Writing the test items. This step begins with determining the number of test items which depends on the type and form of the test. The type of test chosen determines the number of items to be developed. Compared to essay test, objective test requires much more test items. After determining the number of test items, the writing of the test items can be done. The number of test items prepared should be more than those needed in the real test so that when there are some items that do not work well in the development can be dropped without constructing other items. The test items have to be written clearly so as to be easy to understand; and therefore, this avoids misunderstanding. In writing multiple choice items, in addition to write the items clearly, the writing of the choices that consists of the answer key and the distractors has to be similar in content, form, and length. In addition to the writing of the test items, the instructions of what to do and how to do have to be clear.
Step 4: Improving test items. The first draft of the test as the result of writing the test items at the beginning stage still needs a series of improvement. In order to do this improvement, there are at least three steps to follow: (1) read through the whole test, (2) ask other comments, opinions to other people, and (3) try out the draft to the target students. The first step of reading through the draft of the test can be done by the test constructor himself. The second step can be done by asking other teachers or experts to give critical comments, suggestions, and opinions about the test. The third step can be done by trying out the draft of the test to the target students. This last step is used to see the characteristics of the test especially the validity, reliability, practicality and also the effectiveness of the test items like the difficulty index, and discrimination index. These efforts are conducted to improve the draft of the test to make a better test.
Step 5: Final Test Draft. Through the previous steps of developing the test, it is expected that the final test has fulfilled the important aspects as a good test: clear objective, form, and types of the test. The characteristics of the test are also understood: validity, reliability, and practicality. More complete information of the test also includes difficulty index and discrimination index. Answer key for objective test, how to score and grade the students’ answers have to be prepared. For essay test, guidelines to score and grade the students’ answers need to be prepared. Therefore, the final complete test should be clear for students and for the teachers who administer and score the test.
O’Malley and Pierce (1996:17-19) state that there are eight steps to consider in designing authentic assessments: build a team, determine the purpose, specify objectives, conduct staff development, collect sample assessment, adapt existing assessment, try out the assessments, and review the assessments. Each of the eight steps is described as follows. (1) Building a team is described as creating an assessment team of teachers, parents and administrators. (2) Determining the purpose is meant determining the purposes of the assessments for use in planning instruction or for other purposes. (3) Specifing objectives means specifying the instructional objectives to be evaluated with the assessments. (4) Conducting staff development means the staff development for the design team and other teachers on the purposes, use, and development of the assessment. (5) Collecting sample assessment is meant to review the sample assessments for suitability in your school or district to meet local purposes and objectives. (6) Adapting existing assessment is meant where possible, adapting existing instruments to meet local purposes by changing the item content, format, or scoring rubric. (7) Trying out the assessments is meant to give the assessments to students, score the papers, and discuss the assessments with students and other teachers. (8) Reviewing the assessments means discussing the assessments with other members of the team and making the final adjustments to the assessments or the rubrics.
The CL Test Methods
CL Test methods can include tests of all the four language skills of listening, reading, speaking and writing and using a variety of formats (Weir, 1990:42-85).
The formats of listening comprehension tests include multiple-choice questions, short answer questions, and information transfer techniques for testing extensive listening skills; dictation, and listening recall for the testing of intensive listening (Weir, 1990). It is concluded that listening tests, where possible, should cover an authentic performance task. Carroll (1980) suggests the activities that should be covered in the tests of listening. The activities include social interaction, visiting places of interest, travel, medical attention, using the media, and studying for interest. Vallette (1977) suggests short answer questions, dictation including spot/ partial dictation and full dictation, taking notes, and multiple choice questions for the formats of listening comprehension tests.
The formats of tests of reading comprehension include multiple-choice questions, short answer questions, cloze procedure, selective deletion gap filling, C-Tests, Cloze elide, and information transfer. Weir concludes that for testing reading abilities, the use of short answer questions together with selective deletion gap filling are recommended. Focusing on performance tasks in reading tests, the use of information transfer techniques and other restricted response formats are suggested. Carroll (1980) suggests the use of close-ended or restricted-response formats for the tests of reading. Heaton (1989) includes matching, true/ false, multiple choice, rearrangement, and open-ended test formats for reading comprehension test. Madsen (1983) includes editing tests (cloze elide) and restricted response for reading test formats.
The formats for testing speaking according to Weir (1990) include verbal essay, oral presentation, the free interview, the controlled interview, information transfer: description of a picture sequence, information transfer: questions on a single picture, interaction tasks, and role play. Carroll (1980) states that it is difficult to get a single-mode speaking activities. Oral interaction, listening and speaking, is the most commonly-encountered type. So, it is unreal to test an isolated language activity. O’Malley and Pierce (1996) suggest four types of test formats of speaking: oral interview, story or text retelling, projects/ exhibition and experiment/ demonstration. Madsen (1983) includes seven types of test format of speaking: directed response, picture cues, reading aloud, mimicry/ imitation, directed response role play, guided technique, and oral interview.
The formats for testing writing include editing task for indirect methods for assessing linguistic competence; essay tests, controlled writing tasks and summarizing for the direct testing of writing. Carroll (1980) proposes a single-mode writing test, that is, the imaginative essay or the description of a picture, object, and remembered events. Weir (1990) concludes that the concentration of the writing components should be on controlled writing tasks where the specification features of the audience, medium, setting and purpose are stated. Appropriate and adequate scoring criteria and the training of the examiners should be the focus of attention. O’Malley and Pierce (1996) suggest five types of test formats of writing: writing samples, project/ exhibition, experiment/ demonstration, constructed response, and portfolio. Madson (1983) includes seven types of test formats of writing: sentence combining, sentence expansion, sentence reduction tasks, writing a paragraph, writing a narration from a dialog, dictation and free writing. Vallette (1977) suggests many types of test formats of writing, among others are dictation (spot and full), controlled composition, message taking, letter writing, summarizing, writing notes or personal letters, writing business letters, answering an ad, filling out forms.
ENGLISH LANGUAGE PROFICIENCY TEST
This second major section covers the discussion of English language proficiency test including its definition, types, and its development of SELProTest for Indonesian Contexts.
The Definition of English Language Proficiency Test
The term ‘proficiency’ can have three main uses: knowledge or competence, ability to do something specific, and performance (Davies et al., 1999). In the first place, proficiency can be interpreted as “a general type of knowledge or competence in the use of a language”. Secondly, proficiency means ability to do something specific in the language, for example proficiency in English to study in Britain, Australia, the US, etc. Third, the term proficiency is also used to mean performance as measured by a particular testing procedure. Historically, proficiency has been used since 1970s until early 1980’s under the label ‘general language proficiency’ which equals to the term ‘unitary competence hypothesis’. Since then, proficiency has been regarded as multifaceted consisting of several component parts of the performance assessments such as the assessment of speaking or writing skills. Now there is a considerable overlap between the concept of language proficiency and the term communicative competence.
English language Proficiency test, then, is defined as a test of English language which measures how much of English knowledge/competence, ability, and performance someone has learned, or probably more accurately, has mastered. A proficiency test should be differentiated with that of achievement in that a proficiency test should not be based on a particular course of instruction. A candidate is considered to be proficient at the level of novice, intermediate, or superior, in the target language as described in its specific purpose, for example to follow a lecture in Canada, to be a journalist in an English newspaper, to become an English announcer in a television station, etc.
The Types of English Language Proficiency Test
So far, in my view, there are two general types of English language proficiency tests: general English language proficiency test and specific English language proficiency test.
General English Language Proficiency Test
General English language proficiency test is one which measures the test takers’ overall English knowledge, competences, abilities, and performance. This English proficiency could be used to indicate the test takers’ proficiency in joining a program in a target situation which needs general English as one of the requirements in order to perform well in the target situation. Two examples could be presented here to illustrate that statement: Test of English as a Foreign Language (TOEFL), internet-Based TOEFL test (iBT test) and The International English Language Testing System (IELTS).
TOEFL is intended to measure the general English knowledge, competences, abilities, and performance of nonnative speakers of English who wish to study in North American universities and colleges. This test consists of the tests of listening comprehension, structure and written expression, reading comprehension, speaking and writing. The iBT test contains the four language skills: reading, listening, speaking, and writing. The probable format of an iBT TOEFL test consists of 3 passages and 39 questions of reading for 60 minutes, 6 passages and 34 questions of listening for 50 minutes, 6 tasks and 6 questions of speaking for 20 minutes, and 2 tasks and 2 questions of writing for 55 minutes (Phillips, 2006). This test is well-known allover the world and tested in more than 170 countries with the cost of around US$ 15 for each test taker taking the test.
IELTS which was intended to measure the English proficiency of those who want to study in Britain and Australia. Now IELTS is taken by international students who wish to study in the UK, Australia, New Zealand. Canada and even by those who plan to study in the US. As a consequence of this, the IELTS now includes the reading and listening texts that represent the English varieties used in the countries the test takers are likely to study (Taylor, 2006).
Specific English Language Proficiency Test
Specific English language proficiency test is one which measures the test takers’ specific English knowledge, competences, abilities, or performance. This test could be a complete test of general English for specific target needs or a test of specific knowledge, competence, abilities for specific target needs.
A specific English language proficiency test which contains all language skills and components could be developed to measure the test takers’ English proficiency to join a specific intended program, such as to study in a university in Jakarta, to work in the department of foreign affairs, to work in a foreign company, etc. A specific English language proficiency test which contains just one or more knowledge, competence, abilities, or performance could also be developed for a specific target needs. So far, many types of such proficiency test are available, such as TOEIC, Test of Written English, Test of Spoken English. In my view, many other specific English proficiency tests could be developed for any intended target needs.
The Development of SELProTest for Indonesian Contexts
Developing SELProTest for any specific context could be conducted through a specific system. Similar to developing an achievement test, developing SELProTest should at least consist of four steps: specify the English proficiency which the test takers are required to master or perform, determine the types and forms of test, write the test items, and revise the test.
First of all, to develop SELProtest one has to specify the English proficiency which the test takers are required to master. If the proficiency required is just to comprehend reading texts for academic purposes and writing the summary of the texts read, then it is clear that the proficiency the test takers have to master is that they have to be able to write the summary of the reading text for academic purposes they have read. This specific proficiency has to be kept in mind as the basis to develop this proficiency test.
The second step is to determine the types and forms of the test. The types and forms of test have to be in line with the proficiency already decided in the first step. In the case stated in the previous paragraph, whether you like it or not, you have to decide that the test type is writing a summary based on the reading text.
The third step is to write the test items. To write the test items of reading, the first thing that needs to be decided is the reading texts and the types of questions. The number of questions that has to be written is dependent on the time allocated for the test takers to complete the test.
The last step in developing SELProTest is revising or validating the test that has been completely developed. The first validation is conducted by asking other test experts to read through the whole test and provide the necessary comments for the sake of the improvements of the test. The second validation is trying out the test both to the small and then real group of the target test takers. The revision of the SELProTest developed is based on the information obtained from the experts and the try out.
CONCLUSION AND SUGGESTION
To conclude, in developing a specific English proficiency test, a test developer has to know exactly the English proficiency that has to be performed by the test takers. Next, he/she has to determine the types and forms of the test based on the proficiency already determined. After that, he/she has to write the test items based on the types and forms, and revise the test based on the expert judgment and empirical studies.
To end this paper, I suggest that English proficiency tests in Indonesian contexts, or even for other Asian countries’ contexts be self-developed. If this can be realized, we can have more expertise in developing proficiency tests. Moreover, the test takers can spend less money to take an English proficiency test. In order to give more validity proofs the SELProTest developed by Indonesian test developers could be correlated with the existing standardized English proficiency tests like TOEFL and IELTS.
REFERENCES
