Why is there so much variation in the performance of East Asian countries in the skills tested by Ielts?
And do the Chinese need to give up their attachment to traditional teaching in L1 in order to excel? Melanie Butler investigates
Jocelyn Wang, head of teacher development at China’s largest language school chain, New Oriental, recently gave a heartfelt defence of the Chinese system of English language training. At this year’s Iatefl conference, she showed video clips of expert Chinese teachers delivering lectures mostly in Mandarin while the students sat in rapt attention, occasionally taking notes.
This, Wang pointed out in faultless English, was how she had learned the language, and it remained the preferred method for high school students and those preparing to study abroad.
The communicative language teaching approach, in which teachers had been trained at the New Oriental Celta Centre, worked reasonably well with younger learners, she said. But older students and their parents complained when Celta-trained teachers tried to get them to work in pairs or groups, or to practise speaking and writing. This reflected the Chinese tradition of education, shared with much of East Asia, which has made them dominant in maths and science.
So why should the Chinese be forced to adopt British-style communicative language techniques? What was wrong with the old Chinese way?
One way of measuring the effectiveness of a methodology, or indeed of a whole learning culture, is by measuring international test scores and comparing the results of students of other nationalities.
To see what, if any, effect traditional Chinese and East Asian methods of teaching had on outcomes, we analysed the results of the most recent (2015) Ielts Academic English test. We chose Academic Ielts because it is the most widely taken of academic English exams and because it publishes results for the top forty countries and languages, ensuring large samples.
It has a fairly homogenous demographic: academically able 17–25 year olds preparing to study for university, normally abroad, and thus likely to be from the highest socio-economic groups.
Nearly all candidates enrol in test preparation programmes. Looking at the skills profiles for Academic Ielts should show whether traditional Chinese methodology results in lower oral and aural skills, those most commonly focused on in British-style communicative language teaching.
Remember though that in Ielts Academic every single nationality scores lower in writing than in any other skill – which is not the case for General Ielts.
Of course, the differences of scores on a test do not only reflect the differences in learning culture and teaching methodology.
A 2006 Chinese research paper – Kim, M-H & Lee, H-H (2006) Determinants of Toefl scores: a comparison of linguistic and economic factors – showed that for both the paper- and computer-based Toefl tests outcomes were correlated primarily with the degree of language difference between the L1 and L2. The US-administered Toefl exams, which are focused on academic English, did not test speaking but did test language knowledge.
Therefore speakers of Indo-European languages outperform speakers of non-Indo-European languages, with Germanic language speakers outperforming speakers of Slavic or Romance languages.
The second predictor for test success was the amount of English in the environment – the number of native speakers, its use as a second language and the use of English in the media.
To examine the impact of language distance, we started by comparing the 2015 Ielts skills profile of mainland Chinese to those in other Chinese-speaking regions. As Table 1 shows, the skills profile of mainland China is quite unlike that of Singapore and the other Chinese-speaking regions.
All Chinese test-takers score most strongly in the two receptive skills, reading and listening. Both Hong Kong and Singapore have listening as their strongest skill and only students from mainland China score higher on reading than listening. The mainland also shows the lowest difference between speaking and writing, with only 0.1 point between them.
Exposure to English is extremely high in Singapore, where English-medium education is compulsory, and high in Hong Kong, where it is popular.
It is low to medium in Taiwan, where English medium education is rare, although enrolment in private language schools and insistence on native-speaker teachers, largely American and often untrained, is high. So, perhaps low exposure to English is the area to look at.
Mainland China shares similar educational cultures to the other East Asian countries. Of the three largest markets, Korea has low-to-medium amounts of English in the environment, while Japan and Vietnam, like China, have very little.
As we can see from Table 2, however, mainland China still remains at the bottom. Vietnam may benefit from having a romanised alphabet for both reading and writing. And apart from speaking, Japan’s results are the same as China’s. Mainland China is the only place where there is only a 0.1 point difference between the writing score and the speaking score on Academic Ielts, although Vietnam has just 0.2. India, which also has a speaking score 0.2 points above writing, is the only other country to show this profile.
While Chinese speakers outside the mainland generally have stronger language skills than the rest of East Asia, by contrast mainland students have proportionately lower scores in listening and, more especially, speaking than do students from other countries.
According to New Oriental, mainland students object to being made to listen and speak in English, but exposure to spoken English is the very factor which seems to help other Chinese speakers to excel. Traditional teaching delivered in L1 is unlikely to provide the solution to this skills gap. It is up to Jocelyn Wang and her colleagues to come up with a Chinese-style answer.
Natives may not have the answer
Improving the communicative skills of listening and speaking is the main focus of British-style communicative teaching. Enrolment in private language schools is high across East Asia and most parents insist on native-speaker ‘teachers’, so how can New Oriental be right in thinking that Celta-trained teachers are not the answer?
Most teachers in Taiwan, Korea and Japan are Americans, and very few of them have Celtas. Many of them are not trained at all before they arrive, and those that are trained, are likely to follow Krashen’s comprehensible input hypothesis, where learners hear and read simplified language, and are under no pressure to speak. Vietnam, where the UK and Australia seem to be the big players in the market, is an exception.
But if Celta-style communicative teaching is more common there, it is not effective in promoting speaking. Vietnamese comparative scores for speaking are the second lowest of all Ielts markets.
Communicative teaching predominates in Europe and Latin America, but the Ielts scores show that the receptive skills of listening and reading are stronger than the productive ones. Most of the countries which have higher scores for reading than speaking have Latin languages: Columbia, France, Italy and Spain. An Italian teacher told us that reading is easier ‘because written English, especially academic English, uses more Latin words, and the way English is pronounced makes even those words more difficult to follow when you’re listening’.
In fact the countries with the strongest communicative skills have little exposure to communicative teaching or to native speakers. Every country in the Indian subcontinent has higher scores in oral/aural skills, though visa restrictions mean there are few native-speaker teachers. Iran has had no native-speaker teachers for forty years and still scores highest on oral/aural skills. Judging from the evidence from the Ielts results, simply hiring native-speaker teachers with Celtas is not an instant solution to the problem of improving communicative skills.
Pic courtesy: clairemono