International Language Assessment and Local Performance : the case of the Postgraduate Institute of Medicine

Although the world over, English is the premier language in the field of Medicine, in Sri Lanka there is a paucity of studies on English language proficiency in this field. This paper deals with the proficiency of a group of doctors enrolled at the Postgraduate Institute of Medicine, University of Colombo, and investigates their performance on the International English Language Testing System (IELTS), a test of English recognized internationally for studies in Medicine. These doctors were drawn from all faculties of Medicine established in Sri Lanka and also included a small group trained in foreign universities. The paper sets out details of performance at a Pre-Test, and their efforts on an actual version of the IELTS after an eight-month training course. Performance is examined in relation to gender as well as university of graduation. Initially the female candidates demonstrated a higher level of proficiency in certain skills. After training however, this initial advantage was obliterated and the male candidates seemed to be more successful than their female counterparts.


Introduction
It is now widely accepted that English is the international language of science and technology.Graddol (1997) traces the development of English as an international language of science, and points out that during the early years of the twentieth century, the dominant language in this field was German, not English.By the end of the century however, English ruled supreme, and by 1984, the great majority (95%) of publications even by German scientists were in English.The fact that English is the language most relevant for fields of study involving science and technology is accepted by academic communities the world over, and Sri Lanka is no exception.Although in the past, the role of English in academic studies in Sri Lanka had been down-played (Raheem & Ratwatte, 2001), today the importance of this global language has been emphasised as relevant, not only for undergraduate studies but also for those aspiring to university academic posts.In the Sri Lankan academic context, the need for English exists not only for work or study within the university, but also for postgraduate study in institutes overseas.However though this need has been recognised in documents such as circulars issued by the University Grants Commission, and attempts made to cater to these needs through mechanisms such as the establishment of English Language Teaching (ELT) Units in all universities, the issue of English for postgraduate study has not become a particular field for ELT practitioners either as pedagogic practice or as a field for research.This paper therefore attempts to provide some contribution to this area, and is based on a study of performance in English by a group registered at a postgraduate institute.The institute in question is the Postgraduate Institute of Medicine (PGIM) of the University of Colombo, and the focus of the study is a small cohort of doctors registered as trainees at the PGIM.The study examines the performance of these doctors at an international test of English language skills, and investigates this performance in terms of requirements for postgraduate study overseas, gender and university of graduation.

Background to the Study
The PGIM, established in 1980, is the only institute in Sri Lanka responsible for the specialist training of medical doctors.It is a national institute attached to the University of Colombo, and is internationally recognised.Currently all doctors in Sri Lanka who wish to specialise in fields of Medicine are required to complete a postgraduate degree at the PGIM.In addition, such trainees are also required to undergo a period of training in their chosen specialism at a recognised institute overseas.This overseas training is usually obtained in countries such as the United Kingdom, Australia, New Zealand or the United States.To enter these countries for further education, trainees from Sri Lanka have to qualify at a test of English language skills, the International English Language Testing System (IELTS) for Commonwealth countries and the Test of English as a Foreign Language (TOEFL) for the United States.
In the past, the PGIM did not concern itself with these English language tests but in 2003, concerned by the fact that Sri Lankan trainees were finding it difficult to obtain placement because of poor English language skills, the PGIM requested assistance from the Department of Language Studies (LSD) at the Open University of Sri Lanka (OUSL) to deal with this issue.It was decided that LSD would help the doctors to cope with the battery of tests offered as the IELTS, the test accepted by Britain, Australia and New Zealand, the countries that provided training for a majority of Sri Lankan doctors.
IELTS is sponsored by the British Council and owes its genesis to the recognition that many overseas students in Britain do not have the required language ability to cope with academic studies in English.IELTS began in 1964 as the English Proficiency Test Battery (EPTB).This original version was abandoned when it was discovered that the tests were not adequate, and "In the latter half of the 1970s, the British Council was faced with the need to introduce a new.. English proficiency testing system geared to the changes in the needs of sponsored students seeking to come to England." (Seaton, 1981, pp. 121).
A new test was devised, viz: the English Language Testing Service (ELTS) which tested all four language skills and grammar as separate modules.Candidates were scored according to a band scale from 1 to 9, and the overall score was the mean score of all the modules.ELTS was introduced in 1980 and was generally well accepted.However the British Council and the University of Cambridge Local Examination Syndicate (UCLES), who were the joint sponsors of the test commissioned a survey of the test's reliability and validity.The survey conducted by the University of Edinburgh recommended that certain aspects of ELTS be changed.The new version was adopted in 1989 as the International English Language Testing System, and retained the testing of the four language skills in four modules The rating system of 9 bands and an overall score based on the mean was also retained.The grammar component however was deleted (Clapham, 1996).The IELTS test battery in use today thus consists of the following features (Table 1):

The Study
In 2003, LSD began its project at the PGIM to test trainees and check whether they would be able to successfully complete the IELTS test conducted by the British Council, Colombo.The project also included a training component whereby those who were did not reach the required IELTS Band level would be provided guidance to help them cope with the IELTS testing process.Project activities thus included a Pre-Test to provide information on the language skills of the trainees, an eightweek period of language training and a Post-Test to check whether the training had been effective.
In the case of the trainees at the PGIM, it was decided that an abbreviated version of an IELTS test would be administered as the Pre-Test.This abbreviated version involved tests in all four skills but the testing tasks for Listening and Reading were curtailed.For Listening, candidates were required to listen to two passages (instead of three as in the authentic test) and answer only 20 questions.In the authentic test, 40 questions had to be answered.For Reading, only one passage was used (instead of three).The time period for the Pre-Test was also shorter -one and a half hours as contrasted with two and a half hours for the authentic test.The Writing (2 tasks) and Speech components (3 tasks) of the Pre-Test however replicated those of an authentic test.
29 trainees sat the Pre-Test, and on the results of the test, were divided into two groups-one group roughly approximating to an intermediate level of proficiency and the other to a higher intermediate level of proficiency.
A training course was then conducted for 8 weeks (April -May) which dealt with all four skills.At the end of May, a Post-Test was held.This consisted of an authentic IELTS specimen Test, and comprised all the required tasks, the duration of the test being two and a half hours.The skill of Speech however was not tested as performance of the trainees at the Pre-Test had indicated that they were of the necessary standard for this skill.16 trainees sat the Post-Test.
At the end of the Post-Test, a General Knowledge quiz was also administered to the trainees as the training course had revealed a lack of knowledge of world and current affairs, and even a paucity of general information on Sri Lanka.The results of the Pre-Test, the Post-Test and the quiz are discussed below.

Results of Study
In both cases, ie the Pre-Test as well as the Post -Test, the trainees included both sexes and included graduates from all the medical faculties in Sri Lanka.The cohort also included a small number who had obtained their medical degree from foreign universities.This is represented in Figure 1 and Figure 2.
At the Pre -Test, the medical faculties of the different universities were fairly well represented with 9 trainees from Colombo, 7 from Ruhuna, 5 from Jaffna and 3 from Kelaniya.There was only trainee each from Peradeniya and Sri Jayawardenepura.At the Post -Test, the proportions still remained more or less the same with the largest number from Colombo and smaller numbers from the other medical faculties.As the common skills tested at both the Pre -Test and the Post -Test were Reading, Writing and Listening, the study focused on these skills.Table 2 sets out details of performance in these skills and indicates the range of marks and mean per skill.Reading at the Post-Test was identical to that of an authentic test, the marks seem to reflect the fact that the trainees would not be able to cope with the complex and demanding questions set for Reading Comprehension at the IELTS Test.

UNIVERSITIES INVOLVED IN PRE -TEST
These overall marks also need to be set against the required mark for Band 7 rating.Table 3 below sets out the marks required to reach Band 7 for each skill at each Test.It should be noted that there is a discrepancy between the Pre-Test and the Post -Test for marks required for Band 7 rating in the skills of Listening and Reading.This occurred because at the Pre-Test, there were fewer tasks for these skills and hence the Band 7 rating at the Pre-Test had to be adjusted accordingly.These inferences undergo modification when performance is considered gender-wise as set out in Table 4. Table 5 indicates that for all skills as reflected in the mean score for each skill, both at the Pre-Test as well as at the Post-Test, females performed better than the males overall.
For Listening, for example, the female mean at the Pre-Test was 58.63% which surpassed that of the males (52.05%), a phenomenon reflected in all skills.However when compared to the requisite IELTS score, there was similarity between the sexes.At the Pre-Test, to be rated as Band 7, the trainee needed to score 75% for Listening and 67% for Writing.On average, neither the females nor the males reached the required standard for these skills at the Pre-Test.For Reading, however, the females did better than was needed for the mean here was 73.31% while the required Band 7 mark was 68.5%.The males as a group however could not reach the Band 7 rating at the Pre-Test for their mean score was 66.7%, a score slightly below that required.
After training, the gender-wise differences were not so distinct with regard to Band 7 requirements for both females as well as the males were able to score higher than the required standard for Listening and Reading.With regard to Writing however, trainees of both sexes still needed to score a better mark.The initial disparity in skills performance gender-wise as exhibited at the Pre-Test thus was ironed out at the Post -Test.
In terms of range, it could be observed that while at the Pre-Test, the highest score for each skill was recorded by a female, at the Post-Test, males out-performed the females in Listening and Reading, scoring 95% and 90% respectively for these skills.The corresponding highest female scores for these skills were 87.5% for Listening and 82.5% for Reading.The training thus seems to have been beneficial to trainees of both sexes, but male performance, in particular, had improved considerably.
The study also investigated performance university-wise and in Table 5, the performance of the trainees is represented in terms of their university of graduation.The marks given in the table record the mean mark scored by the trainees per university.It needs to be noted that there was only one trainee from Peradeniya and one from Sri Jayawardenepura and hence the mark represented is of one single individual and not the mean.
This Table illustrates graphically that the skill of Writing is weak across all universities.At the Pre-Test, none of the universities registered 67%, the mark for IELTS Band 7 rating for this skill.
Similarly at the Post-Test too, no single university reached the required 68% for Band 7, although the category 'foreign university' came close to this mark, scoring 67%.
These tables set out the universities in rank order, from lowest scorer to highest, for each Test, each table representing a different skill.The configurations in these tables are quite complex and reveal that no trainee from any one university is strong in all skills.For instance, though Colombo scored the second highest marks at both the Pre Test and the Post-Test for Writing, it ranked below Jaffna and Foreign University for Reading while for Listening, it was the third weakest at the Post -Test.The weak performance of the trainee from Peradeniya is well reflected for Peradeniya ranks as the university with the lowest or second lowest marks in all skills.The tables given above thus indicate the language skills of this cohort of trainees, and seem to support that perception that these doctors would not have been able to reach the standards required for IELTS ( at least in two skills) if they had not followed the training.
It also seems worth noting that though the IELTS does not explicitly test general knowledge, the Writing and Speech Tasks require knowledge not of medical specialisms but of the world around us.The inadequacy of knowledge of the trainees vis-a-vis this aspect was highlighted in the General Knowledge Quiz administered after the Post-Test.The quiz was a small test which consisted of three categories of items -(a) World Capitals, (b) Famous Figures and (c) Acronyms.For each category, there were only five questions.Table 9 sets out the performance of the trainees in this quiz.The percentages given reflect the number that got all the items correct for each category.The weakest performance of the trainees is reflected in category 3 where no single trainee was able to get all five acronyms correct.These acronyms were in fact taken from those which had been prominent in the news at the time, and include those that are wellknown such as WTO, SAARC and SLAAS.Very surprisingly only 3 candidates were able to state what SLAAS stood for.Similarly the questions on the world capitals included those of countries such as the United Kingdom, the United States of America, Japan and India.Here too performance was disappointing with candidates identifying Shanghai as the capital of Japan and Mumbai as the capital of India.

Conclusion
This paper sets out the results of an investigation into the performance in English language skills (and general knowledge) by a group of medical personnel.The study is admittedly a preliminary study and although the trainee doctors were drawn from all Sri Lankan medical faculties, the study involved only a very small group at the PGIM.However it seems to highlight certain aspects of the capacity of these doctors in relation to an international language which is of primary importance to the field of Medicine.These aspects are set out in detail below: 1.The performance of the doctors overall at the Pre-Test and subsequently, after training at the Post -Test seems to indicate that the training course had some effect on improving the ability of this group to respond to the tasks set for the IELTS test.This seems especially so in relation to the skills of Listening and Reading.
2. With regard to the skill of Writing however, although, in general, the performance of the sample group had improved as reflected in the scores at the Post-Test, there still was room for improvement as by and large, the trainees had not been able to reach the standard required for postgraduate training abroad.
Given that the skill of Writing is of paramount importance not only in an academic environment but also for professional purposes, this finding needs to be given further attention, and the PGIM should look into ways and means of enhancing the capability of trainees in this skill.
3. The study seems to reflect that the proficiency of the women doctors surpasses that of their male counterparts.This seems to accord with the findings in studies done on gender and proficiency in Sri Lanka (Raheem & Ratwatte, 1996) as well as elsewhere.
4. The study also throws some light on the language ability of these doctors vis-à-vis the university of graduation.The performance of trainees from the Universities of Peradeniya and Ruhuna gives rise to some concern as they seem among the weakest in the group.However it needs to be kept in mind that this cannot be taken as a wholesale indictment as the study only involved those who had volunteered for the training programme.Presumably such volunteers were those who had recognised their own limitations.The more skilled, more proficient graduates of these universities would not have enrolled themselves for the training course, and hence their performance is not reflected in this study.It would however be interesting to examine whether weak performance in English is reflected in or affects performance university-wise in the postgraduate examinations conducted by the PGIM.
5. Notwithstanding the comments given in (4) above, it is disturbing to note that generally speaking, none of the trainees reached the required standard in Writing.This fact needs to be set against the realities of undergraduate medical education in Sri Lanka.All medical faculties in Sri Lanka conduct their degree programmes in the English medium.This implies that all medical graduates have passed through a rigorous five-year long programme which involves academic tasks such as written examination papers, tutorials and assignments, as well as professional medical task in the medium of English.In spite of such undergraduate experience, the trainees have not been able to respond satisfactorily to the complex sophisticated demands of the IELTS test.
This small study of the performance of trainee doctors in English language skills has served to underscore facts not only about the language proficiency of this group, but also of inadequacies in general knowledge.The sample population investigated here are high -calibre students, considered the cream of university undergraduates.They form only a small percent of the trainee population of the PGIM.The facts reflected in this study however give rise to concerns about graduate competence in English language skills.

Figure 2 .
Figure 2. No. of medical graduates involved in Post-test.

Table 1 .
Features of the current IELTS test batterySpecimen copies of the IELTs are available at the British Council which continues to be one of the sponsors of the test, the other sponsors being UCLES and the International Development Program of Australian Universities and Colleges (IDP).The test is thus officially accepted by Australian universities, in addition to being the tool of English proficiency accepted for undergraduate as well as postgraduate study in Britain.For postgraduate study in Medicine, candidates have to obtain a mean rating of Band 7 or more on all skills.
No. of medical graduates involved in Pre-test.

Table 2 .
Range and Mean per Skill (Overall)

Table 2
indicates that the overall mean score in respect of Listening had improved from 54.65% at the Pre-Test to 61.34% at the Post-Test and in Writing from 44.82% to 60.93 %.The overall mean for Reading however had decreased from 70.15% to 64.15%.Since the task for

Table 3 .
Marks required for band 7 rating

Table 4 .
Performance of Trainees -Gender-wise

Table 9 .
Performance on General Knowledge Quiz