Monday, September 15, 2008

Bakri Musa's view on S&M in English

A well thought out piece by Bakri Musa. I got this through email and I don't think Bakri has posted it up on his website yet. I'll link it when its posted online. In the meantime, I've reproduced the article below. It makes reference to a UPSI study on the issue at hand. I'll try to locate a copy of this study and look further into it.

Continue Teaching Science and Mathematics in English

M. Bakri Musa

In May 2003, five months after the government started the teaching of science and mathematics in English in our schools, the Ministry of Education produced a “study” with the incredulous findings of significant improvement in our students’ achievements! All in five months!

Now five years later, research from the Universiti Pendidikan Sultan Idris (UPSI) showed the very opposite results. What gives?

Both studies were prominently and uncritically reported in our mainstream media. That first study was presumably swallowed whole by our policymakers to justify continuing their policy. Rest assured that this second one too would be used for a similar purpose, as an excuse to jettison that same policy.

Despite many attempts I was unable to get a copy of that first study. Nor have I seen it published in any journal, or find any paper credited to its author, raising questions on the credibility of the “study” and competence of its “researcher.”

To the credit of its authors, this later paper is freely available on the Internet, all 153 pages of it. Its lead author is an emeritus professor, a title reserved for retired accomplished scholars, with a dean and deputy dean as his coauthors. Despite its impressive authorship, this study is deeply flawed in its design and conclusions. It does however, expose many weaknesses in the implementation of the policy, in particular the lack of teachers fluent in English.

Embarrassingly Flawed Study

The most glaring deficiency of this second study is its lack of any control group. This is basic in any research design. As the English language policy applies to all schools, you obviously cannot find a control group among current students. You can however find historical control groups by using the test scores of earlier comparable pupils who had been taught and tested in Malay.

With some ingenuity we could still have concurrent control groups, for example, Malaysian pupils attending English schools like Alice School and International School. Another would be adults fluent in English, or even the teachers. If those adults and students in English schools did equally poorly, then clearly the test is not reliable.

When I look at the test questions, it is not only the teachers who are deficient in English, so too are the test makers! Some of the questions are convoluted and would challenge even those fluent in English.

The second flaw is that there is minimal statistical analysis of the data. The pupils were tested and the results simply collated in pages and pages of raw data presented in dull, repetitive and uninformative tables. The authors must be graphically-challenged; they seem to have not heard of pie charts or bar diagrams.

There is also no attempt in delineating the roles of the many variables the researchers have included, like teachers’ English fluency, parents’ educational levels, and pupils’ geographic background (urban versus rural). To do that the data would have to be subjected to more sophisticated statistical analyses, beyond the simple analysis of variance used by the authors. Thus we do not know whether those students’ test scores could be correlated with their parents’ educational levels (a well-acknowledged factor) or teachers’ fluency in English.

There are numerous conclusions based on just simplistic summations of the data, with such statements as X percent of Malay students finding the study of science “easy” compared to Y percent of Chinese or Indians feeling likewise, or R percent of Malay students scoring high versus S percent of their Chinese counterparts. It seems that Malaysian academics, like their politicians, cannot escape the race trap.

These studies were conducted in January, February and July. Even the dumbest students knew that those were not the examination months. They knew those tests “don’t count;” thus skewing the results. The only way to make them take the test seriously would be to incorporate it into their regular examinations.

Besides, in January and February those students had just returned from their long end-of-year holidays during which considerable attrition of knowledge occurred. The difference between the racial groups may have nothing to do with academics but on such extraneous matters as how fast they settle down to their studies.

Of the 27 references cited, there is surprisingly no article from refereed journals. Most (14) are government-sponsored surveys, press releases, and newspaper articles, unusual for a scholarly paper. There are a few books cited, with the most recent published in 2002. There is considerable lag time between what is written in books versus the current state of knowledge. For that you would need journals and attend symposia.

Consequently the researchers’ review on bilingual education is dated. Contrary to their conclusion, it is now accepted that exposing children at a young age to bilingual education confers significant linguistic, cognitive and other advantages. The authors’ recommendation that pupils be taught only in their mother tongue and learn a second language later at a much older age is not supported by modern research.

Studies using functional MRIs (imaging studies) of the brain show that children who are bilingual at an earlier age use their brain more efficiently as compared to those who acquire those skills as adults. For example, when asked to translate between the two languages, “native” bilingual speakers use only one part of their brain while those who are bilingual as an adult use two.

Other cognitive advantages to “native” bilingual speakers include the ability to grasp abstract concepts faster, precisely the intellectual skill helpful in learning mathematics and higher-level science. The higher scores for non-Malays may well be the consequence of their earlier and more extensive exposure to bilingualism than Malays.

Revealing Findings

The study nonetheless reveals many useful findings. I fear however, that these nuggets of information would be lost by those who care only for the study’s unjustified conclusion to discontinue the present policy and revert to teaching science and mathematics in Malay. That would be a retrogressive step.

This study is only a snapshot; it does not enlighten us as to trend. It could be that the results would continue to improve. It is thus presumptuous for the authors to make a sweeping policy recommendation based only a limited snapshot study, and a poorly-designed one at that.

UPSI in its previous incarnation as Sultan Idris Teachers’ College was a hotbed of Malay nationalism. This study is less an academic research and more political polemic camouflaged as a pseudo-scientific study to justify its authors’ nationalist bias. Their data and methodology just do not support their conclusion.

The study found that fewer than 15 percent of the teachers were fluent in English, and that most teach using a combination of both languages. That is putting it politely. In reality they use bastardized or “pidgin” English. If those teachers lack English language skills, how could they teach any subject in that language? The fault here is not with the policy, rather its implementation. We should first train the teachers.

In its naivety the government spent over RM3 billion to equip these teachers with computers, LCDs and “teaching modules” to help them in the classroom. Many of those computers are now conveniently “stolen,” plugged with viruses, or simply left to gather dust as those teachers lack the skills to use them effectively.

The only beneficiaries of that program were UMNO operatives who secured those lucrative contracts. Had the government spent those precious funds to hire new teachers fluent in English, our students would have been better served, and the policy more effectively implemented.

This study missed a splendid opportunity to find out what those students, parents and teachers felt about the policy. It was as if those researchers and their field workers (undergraduates in education and thus our future teachers) were interested only in administering those tests, collecting their data, and then getting back to campus.

Surely those parents and teachers had something to say on the policy. What do the teachers feel about the billions spent on computers? Are they eager to learn and teach in English or do they harbor nationalist sentiments and resent the policy? Those surveys would have helped considerably towards implementing the policy better.

A Better Way

I support the teaching of science and mathematics in English. I go further and would have half the subjects in our national schools be taught in English, including Islamic Studies. The objective should be to produce thoroughly or “native” bilingual graduates, able to read, write and even dream in Malay and English. That is the only way to make our graduates competitive.

I put forth my ideas on achieving this in my earlier (2003) book, An Education System Worthy of Malaysia. I would start small, restricting the program to our residential schools where the students are smarter, teachers better, and facilities superior. Work out the kinks there first, only then expand the program.

I would also convert some teachers’ colleges into exclusively English-medium institutions to train future teachers of English, science, and mathematics.

In rural areas where the level of English in the schools and community is low, I would bring back the old English-medium schools, but modifying it significantly with pupils taught exclusively in English for the first four years (“total immersion”). Malay would be introduced only in Year V, and only as one subject.

Since Malay would not be taught in the first few years and only a limited subject later on, admission to such schools would be restricted only to those with already near-native fluency in Malay or whose habitual language is Malay. Further, such schools would be set up only where the background level of Malay in the community is high, essentially only in the kampongs.

If we were to do otherwise, as having such schools in the cities where the level of English in the community is high and Malay low, those graduates would not be fluent in our national language, as during colonial days. It would not be in the national interest to repeat that mistake. Besides, the problem of our students’ deficiency in English is most acute in rural areas. Thus it makes sense to establish English-medium schools there.

There are many challenges to the policy of teaching science and mathematics in English. One thing is certain. We will never resolve them if we listen to ambitious politicians playing to the gallery or rely on less-than-rigorous “researches.”

4 comments:

Pragalath said...

Can I have a copy of the report once you have it? I would like to add it on digitalibrary.my/

WY said...

Kian Ming,

I hope Bakri's sound argument would sway you from KJ feeble statements..

Anonymous said...

Are you aware of what S&M means in most varieties of English (as well as many other languages)??

Anonymous said...

Obviously, the overall issue is the abysmal state of the public-run educational system in Malaysia.
From time to time, I receive snippets claiming that children of certain ministers are attending local private-elite schools as well as boarding schools abroad. Is there any authoritative source for such information? Seems a worthwhile effort for those who have the resource to conduct such an exercise and put this information in the public domain. If it turns out to be true that a significant number ofchildren of the people responsible for our education system, are in fact going to "better" alternative school systems, this would be a real eye-opener particularly to those who are either ignorant or blind to the sad state of our public educational system.