27 Jan 2011

New Blood

If the escalating impact factors are due to increasing number of researchers, where do they come from?

Western politicians love to blame Chinese workers for everything. But this time they are not too far off. The last decade saw a rapid increase in research papers from developing countries. In fact, China is already second (after USA) in the number of biomedical papers produced last year. Although the quality of many papers still has room for improvement, more and more top scientific papers are coming from China.

There are many reasons behind the rise of developing countries. Many countries enjoyed substantial economic growth and could support scientific research. Globalization and easy access to scientific articles online also stimulate scientific development.

I found a new reason during our recent visit to China.

A medical professor told us that not only academics but also most doctors in China had to publish papers. It is almost impossible to get promoted to a consultant post in urban hospitals without publications. In the past, publications in Chinese journals were still acceptable. Nowadays, administrators are more ambitious and ask for papers in international journals. Worse still, many international journals do not accept case reports any more. So original research works are required.

While I believe all medical doctors should be trained in evidence-based medicine and be able to critique biomedical works, asking everyone to write papers appears to be pushing too far.

“Do you mean,” I found it difficult to understand, “that even doctors in small service hospitals have to do studies and publish papers?”

“Yes, indeed.”

“How do they get the time and funding for research?”

“Well, they don’t have any.”

“That is forcing them to …” I was forced to conclude. And we fell back in silence.

20 Jan 2011


What did I learn from that list?

To understand the list, we first need to know how scientific journals are ranked. Journals in different categories are ranked according to the impact factor. For example, the impact factor of a journal in 2009 is calculated as the number of times articles published in 2007 and 2008 were cited during 2009, divided by the total number of articles published by that journal in 2007 and 2008. Simply said, it reflects the average citations per article a journal receives.

There are many ways to boost the impact factor. Some journals reduce the number of original articles and increase the number of review articles, which are known to be cited more often. Journals affiliated to certain societies have particular advantage because they can publish guidelines. I have also encountered journals that would explicitly ask authors to cite their articles before the papers would be accepted.

Let’s for the time being ignore the differences among individual journals’ strategies. The list is enlightening.

Overall, the impact factors of most journals are on the rise. This does not necessarily mean that new papers are more important than old ones. Rather, the phenomenon can be explained by the rapid increase in the number of journals in each category and the number of researchers. In fact, journals whose impact factors did not increase significantly in the last decade almost invariably fell back in rank.

Secondly, there was little change in the ranking of top journals in each category. This is due to the submission pattern of researchers. It is natural for researchers to submit their best works to the top journals. The other journals could just receive less important works and even papers that have been rejected by top journals. Thus the difference between top journals and the other journals perpetuates.

Moreover, the increase in impact factor and number of citations did not distribute evenly among all journals. The increase was more in general journals than in highly specialized journals. Although specialized journals may enjoy good years when the field is rapidly advancing, the downturn can be equally harsh. On the other hand, general journals can easily turn to more promising fields when the environment changes. This, of course, should apply to other industries as well.

13 Jan 2011

Time Capsule

Last week, Szeto showed us the list of stocks he bought in 2001-2002. [See http://ccszeto.blogspot.com/2011/01/list.html.] I could only be jealous. Incidentally, I went through another electronic time capsule at the same time. It was the list of medical journals 10 years ago.

6 Jan 2011


Mark Zuckerberg, the founder of Facebook, is TIME’s Person of the Year 2010.

When one uses the computer everyday, it is easy to take the way of work for granted. The change in the last decade, in fact, is astonishing.

When I was a medical student, people were still searching gigantic directories for reference articles. Just before I graduated, it became possible to do computer search. Even so, the database was not web-based. Instead, the library received a CD ROM containing the information of scientific articles every 3 months. One could only know the latest publications by reading many journals or attending medical conferences. Nowadays, with a click on my keyboard, I can find not only articles published this week but also those that will be published in the coming months. As for medical conferences, I can download the presentation slides, or even watch the video of the presentation online.

In the past, when people submitted papers to journals, they had to make several photocopies of the manuscript and send them by express mail. The famous DNA paper by Watson and Crick in Nature, for example, was prepared by James Watson’s sister with a typewriter under his constant encouragement: “You are participating in perhaps the most famous event in biology since Darwin’s book!” With mails to and fro, the review process usually took several months. You may think this kind of ancient communication must occur before the 1990s. In this you are wrong. My last paper published this way was in 2004.

I still remember vividly the moments when I received reply letters from medical journals. Is my paper accepted? The anticipation and apprehension were no less than those from love letters. Now all we get are e-mail notifications. Certainly time has changed. When young lovers break up nowadays, they just send each other text messages by phone. Worse still, some simply change their status on Facebook. I wonder when we will post scientific papers on Facebook and be evaluated by the number of ‘likes’ from readers.

Maybe we should not complain. Under the current research assessment exercise, academics should submit as many papers as the phone numbers a bachelor gets at dating parties. As such, we should settle with Facebook. Or perhaps Twitter?