A Four-Step Approach for Improved Syllabus Design Coherence

Paul Moritoshi

Sanyo Gakuen University

In such situations, teachers might opt to test for the extent to which students have fulfilled the original course aim, i.e. that which the syllabus as a whole was intended to achieve. Such a test might generally be considered a proficiency test, depending on what that aim is. Alternatively teachers may decide to test for the extent to which students have acquired the course content, i.e. the degree to which they have learnt or understood the material covered. This would be an achievement test. The two are subtly different: the latter assesses students knowledge, while the former generally assesses students ability to apply that knowledge. However, in a divergent program of study, it would be very difficult to design a single instrument to test for both aim and program content uptake because they are now too disparate.

Figure 1: The effect on assessment of a divergent program of study

It would be more desirable to use an approach to syllabus design which, from the outset, makes strong, clear, well reasoned, and informed links between these three components of syllabus specification. One possible approach might be to work through the whole process backwards, i.e. address certain issues relating to test design first, then go on to specify the aim and design the program of study. This approach helped to overcome just such a divergence in my own university level conversation courses. By answering four questions taken from Bachman and Palmer (1996), designers supply themselves with the information they need to produce syllabus aim, program of study and test specifications which are much more closely linked and focused. This paper describes how to work through this process, using my conversation courses as a contextual example, and highlights the potential advantages which this approach has to offer.

The Four-Step Process

The consensus in the literature relating to teaching and testing seems to accept that a test should relate in some way to the course which precedes it (Oller, 1979; Brindley, 1989; Stern, 1992; McNamara, 2000). The first step is therefore to firmly link the tests purpose to the courses aim. In the case of my English conversation courses, the tests purpose was to measure the extent to which examinees could participate actively in a one-to-one conversation in English. Therefore, the course aim was to facilitate development of students English language and conversational skills, to enable them to participate actively in English conversations.

The second step is to describe what Bachman and Palmer term the Target Language Use (TLU) domain, i.e. the situation or context in which the test taker will be using the language outside of the test itself (1996, p. 18). The resulting statement facilitates the design of courses which account for that domain from the outset. Since students in my conversation courses were encouraged to participate in homestays overseas, particularly while studying at the universitys sister institution in New Zealand, this was my TLU domain, albeit a rather vague one.

In the third step, the course designer should identify and exclusively define the construct to be tested. While the first two steps are relatively quick and easy, an informed solution here may require background reading. My construct definition for conversation was a relatively informal (Cook, 1989), cooperative, purposeful interaction in real time (Brazil, 1992), being relatively unrestrained in topic (Coulthard, 1985) and having unplannedness . . . unpredictability of sequence and outcome, potentially equal distribution of rights and duties in talk (Lier, 1989, p. 495) and spontaneity (Cook, 1989, p. 116).

The fourth and final step is to decide what tasks and topical content the test should include. The closer these match the activities and subject matter found in the actual TLU domain the better (Bachman & Palmer, 1996). These tasks and content then act as the basis for the program of study. To continue my conversation example, the tasks used were those behaviours considered necessary for proficient conversation: greeting, initiating and changing topics, questioning, and closure. Content included topics commonly used in homestay settings: weekend or vacation plans, travel experiences, family, friends, pets, food, home life, and explanation of the Japanese way of life and culture.

To maintain the tests content validity, test task and content selection must, where practicable, be proportionally representative of the TLU domain itself. These relative frequencies can also inform the amount of instruction time allotted to these items in the program of study, though these allocations also need to account for their relative importance and difficulty. To refer back to my conversation course example one last time, turn taking, questioning, backchannelling and nonverbal communication are commonplace, while topic changing occurs less often and greeting and closure occur only once per conversation.

In reality, though the four-step process is performed in roughly the order presented here, the problems they present are often considered concurrently, since one may impinge upon others (Bachman & Palmer, 1996).

Why Use This Approach?

How can this approach help us as syllabus designers/teachers? Perhaps the clearest way to demonstrate is to examine the problems encountered during my first semester conversation courses and then show how the four-step approach helped to overcome them in the second semester. It is important to note first that the two other basic components of my syllabi, i.e. the aim and the test, were the same in both cases. Only my program of study designs, acting as a bridge between these two terminal elements, were changed.

My first semesters programs were based on what might be called an intuitive approach. They were derived almost exclusively from popular conversation coursebooks, selected because they contained the kinds of topics, language items, and activity types which seemed to work towards my aim. The activity designs also seemed to be student-friendly by offering the right amount of support via model dialogues. Each program of study was created simply and quickly by superimposing the associated coursebooks table of contents onto the lesson schedule. Generally, the sequencing followed the order in which the units were presented in the coursebook. Conversely, having already worked through the four steps described above in order to design the first semester tests, my second semester conversation program designs were derived through this approach.

Syllabus coherence

By the end of the first semester courses, the situation was identical to that shown in Figure 1 above. Though intuition had given me a good starting point (the aim), it had not helped me to select materials, activities, nor a methodology which would realise that aim. Those selected were, in short, counterproductive.

Conversely, the four-step approach provided a considered and informed rationale. Step 1 had anchored the two ends of the syllabus (aim specification and test design) firmly in place, then steps 2 to 4 had bridged the gap between them. Step 2 described the TLU domain, which provided a context for the program of study. Step 3 exclusively defined the construct under study so that all concerned knew exactly what the course would focus on. Step 4 generated the program contents which would also form the basis of the test design. The three components of syllabus design were therefore strongly linked from the start, providing a highly focused program of study. This more desirable situation is illustrated in Figure 2.

Figure 2.: The effect on assessment of a coherent program of study

It was now possible to test for both the aim (proficiency) and content acquisition (achievement) and to provide much more specific, useful feedback to students. A further advantage of such informative feedback is that it can provide the basis for aim specification for subsequent courses (Figure 2), which the rudimentary feedback from the first semesters tests could not do so well. Though some might argue that a proficiency test and an achievement test cannot possibly co-exist in a single assessment tool, evidently, where the course content relates to those items which make a person proficient in, for example, conversation, it is entirely possible.

Materials dependency

Real life conversations require non-native speakers to use whatever conversational strategies and English language resources they have acquired, in a flexible, spontaneous way. However, my first semester program designs and the materials on which they were based, did not help students to develop these capacities. During practice activities, students would just work through a model dialogue verbatim, filling the gaps verbally as they went. When they had finished the dialogue as presented, they would put their coursebooks down and start speaking, off topic, in their native language.

Even in the free conversation exchanges, intended to give students the opportunity to use English without restriction, little or no attempt was made to go beyond the model dialogues. Students generally didnt extend the conversation, volunteer information, change the topic, ask follow-up questions, engage in spontaneous turn taking, or close the conversation.

On the rare occasions when these essential conversational behaviours did occur, it was usually in Japanese, not English. To make things even worse, students had their heads down while reading the models, which severely limited the eye contact necessary during conversation for signalling turn changing and gauging reactions. Students dependence upon the materials and/or their unwillingness to go beyond them virtually eliminated any need to use, let alone develop, their conversational strategies and English language.

Although a full explanation of this phenomenon is beyond the purview of this article, it may be explained in part at least, by the over-reliance, in junior (and senior?) high school EFL classes, upon a method which reinforces students beliefs that they need only practice what is presented: the Presentation-Practice-Production method.

The four-step approach generated a clear construct and a list of essential conversational behaviours on which to focus. My second semester programs explained and exemplified these in everyday terms and used activities which targeted them. For example, the activity Poker Face had two students conducting a 1 minute conversation. Prior to the conversation, one student from each pair was secretly instructed not to give any kind of reaction to what her partner says. This experience helped to introduce and explain the idea of giving a reaction (backchannelling), and showed its importance during conversation.

As proficient conversationalists in their native language, students already had an innate understanding of such behaviours. However, by highlighting them overtly in this way, students now understand why reading example dialogues verbatim does not constitute genuine conversation. The change in mindset took time, but students are now much more willing, and able, to participate in genuinely communicative English conversation. They no longer depend upon models. Indeed, example dialogues are now absent from the new material and activity designs. Those currently in use require students to use the conversational strategies and English language that the course set out to promote. In brief, students no longer talk at their partners, they talk to them and with them, a distinction clearly outlined by Brazil (1992).

Content relevance

The coursebooks selected dealt with topics such as showing a foreign visitor around Kyoto, jobs, and money. The activities they contained often required students to role play the part of fictitious coursebook characters. However, these topics and activities generally failed to generate much interest. During practice, students just seemed to be going through the motions. When asked why, it was found that they felt they were unlikely ever to have conversations in English on such topics and they couldnt relate to the characters at all. The content was perceived as irrelevant.

Specifying the TLU domain, tasks and topical content at the start of the second semester courses provided students with an explanation, not just of what the content would be, but why it was relevant. Students were better able to understand the contexts in which they are likely to use the language items and/or skills included on the course and to understand the relevance of the course content to that domain. Also, with the new material and activity designs, students had far more opportunities to talk about themselves and the things that interest them. Informal feedback from students at the end of the second semester indicated that perceived relevance of course content had significantly improved.

Professionalism, responsibility, and innovation

The intuitive approach essentially handed responsibility for many of the syllabus design decisions to the coursebook authors: decisions relating to program content, activity types, material design, and methodology. Such practice has been strongly criticised as leading to a degenerate syllabus (Sinclair & Renouf, 1988, p. 146) because in such cases, programs are not founded upon an understanding of language, language acquisition, language teaching, learning, or use, but on an analysis of other coursebooks and pedagogic grammars (Willis, 2000, p. 3).

The four-step approach encourages attaining a higher level of professionalism and responsibility for the programs we design and implement. It requires syllabus designers/teachers to use their own creativity to make and utilise a more diverse range of materials and activities. For example, after brainstorming the conversational behaviour of nonverbal communication to elicit various types and examples, I used a Mr. Bean video to demonstrate its communicative potential. When a syllabus is enslaved to a coursebook, the need for such innovations is reduced.

Conclusion

This paper has outlined an approach to syllabus design which requires designers to specify four aspects of the test they will use at the end of the course: test purpose; Target Language Use domain; construct definition; and test task and content. These specifications are then fed back into the program of study to create a syllabus which strongly links aim, program, and test as a coherent whole. The paper has also shown how such an approach is preferable to one based on intuitive selection of, and over-reliance upon, coursebook materials. Though intuition can often be a powerful and useful tool, it can apparently also lead one astray unless informed by a wider range of sources.

No hard data can be provided to show that the four-step approach to course design was more efficacious in my course than the intuitive approach which preceded it. However, the type of program of study which the four-step process naturally tends to produce is more systematic, raises awareness of genre-specific rules, strategies and pitfalls, and increases sensitivity to the underlying processes (Dornyei & Thurrell, 1994). These outcomes are all consistent with my original aim.

Also, informal feedback from students suggests that the second semester courses were more enjoyable, that the topics more fun and interesting, and that the instruction and practice had engaged students to a higher level. Though some did express a preference for continued use of coursebooks, they did not voice dissatisfaction with the content used. Finally, by the end of the second semester, students were visibly less hesitant and more confident and capable during conversation practice.

The four-step process seems to offer a viable and systematic approach to syllabus design for those with the freedom to specify their own course aims, programs, and tests. The contextual example provided has shown that its application can greatly enhance the internal consistency of a genre-specific syllabus. It might also be useful to those with little or no experience in designing their own courses, or to those who have to design courses in areas of study with which they are unfamiliar. Its application may also extend to designing other genre-specific courses -- for example speech and debate -- and may not be limited to spoken discourse.

References

Bachman, L., & Palmer, A. (1996). Language testing in practice. Oxford: Oxford University Press.

Brazil, D. (1992, January). Speaking English or talking to people. Lecture given at Sophia University, Tokyo.

Brindley, G. (1989). Assessing achievement in the learner-centred curriculum. Sydney: Macquarie University, National Centre for English Language Teaching and Research.

Brown, J., & Hudson, T. (1998). The alternatives in language assessment. TESOL Quarterly, 32 (4), 653-675.

Cook, G. (1989). Discourse. Oxford: Oxford University Press.

Coulthard, M. (1985). An introduction to discourse analysis. New York: Longman.

Dornyei, Z., & Thurrell, S. (1994). Teaching conversational skills intensively: course content and rationale. ELT Journal, 48 (1), 40-49.

Hughes, A. (1989). Testing for language teachers. Cambridge: Cambridge University Press.

Lier, L. van (1989). Reeling, writhing, drawling, stretching, and fainting in coils: Oral proficiency interviews as conversation. TESOL Quarterly, 23 (3), 489-508.

McNamara, T. (2000). Language testing. Oxford: Oxford University Press.

Oller, J. (1979). Language tests at school. New York: Longman.

Sinclair, J., & Renouf, A. (1988). A lexical syllabus for language learning. In R. Carter & M. McCarthy (Eds.), Vocabulary and language teaching (pp. 140-160). London: Longman.

Stern, H. (1992). Issues and options in language teaching. Oxford: Oxford University Press.

Willis, D. (2000). Syllabus and materials. Birmingham: University of Birmingham, Centre for English Language Studies.