29 Apr 2005

How is Singapore Science Really Doing?

By Acidflask...
 Posted by Hello


The Singapore government has been lauding Singapore as a technologically advanced nation and a world-class leader in research and development, especially in recent years with the earmarking of billions of dollars of taxpayers' funding for encouraging scientific research, particularly in the life sciences. Yet a closer inspection at scientific statistics tells a different tale from the official party line.

The statistics in question are Thomson ISI 's Essential Science Indicators. ISI publishes the Web of Knowledge , one of the world's most important citation databases. (One can think of it as a giant database of academic trackbacks.) Many scientific journals that are deemed of suitable quality for inclusion are indexed back to 1945, the year after WWII that saw a surge in research ouput. Conventional wisdom claims a strong correlation between the number of citations and the relative importance of a paper in that particular field, particularly when compared to the field's average. Hence in academia, where collaboration and building upon the work of predecessors (and the "shoulders of giants ") is critical for progress, citations are one of the most powerful forms of evaluation in terms of what peers think of one's work. ISI's database, being the largest, is therefore arguably one of the most reliable sources of information.

In terms of ISI's citation statistics, Singaporean academia sucks big time. In the years 1999-2003, Singapore scored below average in terms of impact factors in 19 out of 21 fields tracked by Thomson ISI. The only above-average disciplines were mathematics (10% above the global average) and oddly enough, agricultural sciences ( +48 ). In contrast, engineering and physical sciences, which Singapore boasts as its strengths, perform miserably: materials science ( -5 ), chemistry ( -11 ), computer science( -18 ), engineering ( -18 ), physics ( -38 ), and geosciences ( -52 ) were all below average. This contrasts with statistics suggesting prolific scientific output in computer science, engineering, materials science, physics, and mathematics in particular, in terms of sheer numbers of papers published.

Even in the life sciences, the skew toward trendy disciplines is obvious. Consider these two categories and judge for yourself:
hot: agricultural sciences ( +48 ), plant and animal sciences ( -2 ), microbiology ( -5 ), biology and biochemistry ( -14 ), molecular biology ( -19 )
cold: immunology ( -33 ), neurosciences (-36), clinical medicine ( -36 ), ecology/environmental ( -37 ), pharmacology ( -40 ), psychology/psychiatry ( -50 )

Although I am no expert in the life sciences, the preceding dichotomy suggests that pumping money into the life sciences has developed a chasm between the "hot" (easily commercializable) life sciences and the "cold" (traditional and/or non-commercializable) life sciences.

One comment on the methodology: scientific contributions are attributed to Singapore if at least one author's address is reported to be in Singapore. Thus the data do not discriminate between the output of true-blue Singaporean researchers and the output of luminaries who were encouraged financially to grace Singapore with their presence but otherwise have no interests in Singapore as a nation-state. Note also that money is being pumped preferentially into disciplines that happen to have the highest average citations in the field, in particular molecular biology. Considering the bureaucrat's affinity for numbers, is there some statistical gerrymandering to pull up Singapore's ranking in citations statistics?

Most worryingly, clinical medicine ( -36 ) and education ( -54 ) both fare extremely badly in the citations game. Most academics believe that good teaching goes hand-in-hand with good research, because only a good educator will be able to practice what he preaches and only a good researcher, current with the latest happenings in the field, will be able to provide the most relevant education. The connection between good researcher and good doctor is even more obvious. So given the absymal relative lack of research, how is it possible to maintain our "world-class" education and health-care systems? Or unless one is willing to stretch the idea of being "world-class" to include even being far below the world average, were they even "world-class" to begin with?

The conclusion is crystal-clear: we are very good at producing unstimulating scientific research. In other words, we are good at wasting trees. A conclusion that is bolstered by considering that in the years 1993 - 2003, Singapore was ranked 36th in prolificity (i.e. number of papers published in all disciplines) out of 145 countries, but only 92nd in terms of relative impact (citations per paper).

Apologists for the current régime are quick to invoke the usual arguments that Singapore has limited human resources and no natural resources, and is also starting relatively late in the game, and hence such comparisons are biased against Singapore and should not be taken seriously. But comparing Singapore's statistics with other countries quickly debunks such entrenched self-pitying beliefs. Israel, for example, has a population of slightly over 6 million people. It has arguably the fewest natural resources of any Middle-Eastern country, and on top of that has been embroiled in strife and violence since time immemorial. By the preceding arguments, Israel should be one of the world's most ill-developed countries technologically, if for nothing else for having to spend huge gobs of resources just to survive, yet it is not. The citation statistics back up Israel's reputation for high technology: it is ranked 16th in productiveness and 24th in relative impact. South Korea is now by far the most scientifically prolific of the Five Asian Tigers. Belgium , a relatively small country with only 10 million people, has 9.5 papers per 1000 capita and 9.43 citations per paper (ranked 19th). In comparison, Singapore's statistics are 8.6 papers per 1000 capita and 4.58 citations per paper. Even Lebanon , still recuperating from a 16-year civil war that only ended in 1991, managed to beat Singapore in terms of relative impact of papers in computer science (1.12 v. 1.03), a field which Singapore considers among its greatest strengths. The consensus across various statistics speaks for itself.

In conclusion, Singapore's quest for world domination in the scientific sphere must address the grim reality painted by this (admittedly limited) analysis:
That the top 20 countries are leaps and bounds ahead of Singapore, if for nothing else for the sheer availability of resources. So by definition it will be very difficult for Singapore to match up against these truly world-class countries.
We can write many papers, but obviously not many people care about them.
Singapore is not a major contributor to the physical sciences, even in per capita and per paper terms. Even in the life sciences, she is not doing well consistently across the board.
Comparisons with similarly-sized countries show that we are simply not holding the fort even in our own middleweight class.

If this is what a quick look at a few numbers can reveal, how many more ugly truths need to be dug up from more in-depth analyses before Singapore will heed this sobering wake-up call?

3 comments:

Anonymous said...

Great work!
[url=http://zccotgoo.com/gzqc/ngcl.html]My homepage[/url] | [url=http://raoxzxch.com/ivlu/ckrz.html]Cool site[/url]

Anonymous said...

Great work!
[url=http://zccotgoo.com/gzqc/ngcl.html]My homepage[/url] | [url=http://raoxzxch.com/ivlu/ckrz.html]Cool site[/url]

Anonymous said...

Well done!
http://zccotgoo.com/gzqc/ngcl.html | http://sdbonomy.com/wlcj/yuoi.html