Science is higher than suspicion

We publish a new article by a candidate of biological sciences, an employee of the Zoological Museum of Moscow State University George Lyubarsky. His publication “Reforming Russian science: from global to private” (TRV 9 (28) dated May 12, 2009, p. 4-5) attracted great attention to the readers of our newspaper.

I would like to think that all kinds of political games, economic passions and personal ambitions are not related to science. On the other hand, people around come across more and more experienced ones, up to cynicism, and everyone knows that everything is bought, everything is sold and stupid to look in the subwine world of the Absolute. Others even say with meaning - ". And above ”, but now this is not about that.

Surprisingly, many people who are very interested in the state of affairs in science do not know about the mechanisms of its functioning. In addition to all kinds of fights, intrigues, scandals and other everyday joys, there are guidelines that allow you to understand how it works. These are, for example, scientific indices: impact factor, citation index and Hirsch index.

Impact factor (if) is a numerical indicator of the importance of a scientific journal. It is considered very simple: the number of references to articles from the magazine in other magazines, attributed to the total number of articles in the journal. Of course, there are subtleties. For example, the index is considered for a certain period, for two years. If someone begins to quote after 30 years-it does not fall into the index, although the article was clearly non-trivial. Another subtlety is a list of magazines included in the calculation of the index. In total, 8400 magazines from 60 countries are indexed, but there are much more scientific journals in the world, not all are included, but the best are guaranteed. And the average too. In addition, only quoting in articles, and not in letters to the editor, notes, reviews and other scientific small periodicals, are considered. The impact factor considers Thomson Scientific (former Institute for Scientific Information-ISI), annually the results are published in the Journal Citation Report.

The same Taxon institute considers the citation index (SCI)-the number of references to this article of some author. There are several types of citation index, in different fields of knowledge. You can see a lot by simply starting the search on sites http://scholar.google.com/ and http://quadsearch.csd.auth.gr/index.php?s=2&lan=1 . Counting methods

There are quite a lot of citation index for this article - it depends on which selection to include it and how many publications to take into account.

To find out a similar characteristic for the author, and not for a separate article, they are often used by H-index-Hirsch index (see http://arxiv.org/abs/physics/0508025 ). This is a characteristic of productivity, which reflects the number of the most cited articles by the author and their level of citation. The Hirsch index is needed to rely on a thinner indicator than just the total number of articles (which no one reads?) Or the total number of links (his friend, whom he himself refers?).

If you arrange the author’s articles in the quotation order, you get a list - from the most quoted to non -to -taxed. The Hirsch index is the freight number of the last article in this list, the number of links to which is not less than this serial number. For example, on the 15th citation article-19 links, on the 16th-12 links. So, the Hirsch index in this example is 15 (very decent). And if this index is less than 10, there is little chance of snatch a good scientific position. Hirsch (and then others) in his article showed that this factor correlates with recognized “success” indicators.

All indices are affected by the customs of this science. For example, doctors quote each other more willingly and more abundant, and therefore the indices of their magazines are higher than mathematical and H-itxes of doctors are slightly higher than in mathematicians. So between the sciences the index works poorly, in general it depends on the life expectancy of the scientist: in general, the longer you live, the more you published and the more links to you. It’s bad to die the young: the “Hirsch” by Evarist Galois was forever remained 2. Large scientists now have a Hirsch index, equal to 20-30-50 and higher. Some champions have in general for 100. There is a database www.slac.stanford.edu/spires/ where you can see such things for physicists.

Why are all these remarkably simple indicators? They do not reflect the qualities of the work of scientists and magazines. Almost quite in general they are not displayed. Those who are engaged in the same field of knowledge can judge the quality of the work of the scientist. And they stumble upon his mistakes with their own non -armed head or feel how his work helps them.

Why are indexes needed? Of course, they do not need a scientist at all, although it is before him that they can boast of their “Hirsch”. Those who actually know what the scientist is worth, indexes are not needed. They are needed by those who do not know and have no opportunity to find out. It is necessary to look at 150 resumes, and then time ... And then indexes help. In which magazines is this scientist printed with which impact factors? And what is his Hirsch? Yeah. It is approximately clear who it is and what is worth it.

According to these indices, candidates for one or another place are evaluated, career issues are resolved, taken into account when distributing grants and in general - they are guided in the stormy worldly sea in the conditions of a lack of time to receive quality information. Hirsch-index does not replace the expert assessment of competent colleagues-on the contrary, it is a weak similarity of this assessment. For example, the level of a good candidate of sciences or, for example, postdrod in the natural sciences is about 5 in Hirsch. This does not mean that everyone with the five is good. And that someone with the seven is better. The index only measures what it measures is the number of links in such conditions. But you can, if you do not know anything, take into account this.

Here we need to pause and fix: in science there are “objective measures”, universal, worldwide methods for evaluating the work of scientific journals, individual scientists and entire countries. In fact, after all, Hirsch-index can be calculated for an entire country, and already calculated: http://opec.ru/docs.aspx?id=385&ob_no=86405 .

The question arises: is this data it is possible to “wind up”, can it be somehow so that these indicators are more favorable? For the author, magazine or country as a whole.

Of course, measures themselves beg. You can refer to friends so that they refer to you (self -digitalization in these indices are usually not taken into account). And you can try to create your own index, which will be more favorable to take into account. Indeed, each index has its own details and features - a slightly different base of the articles with which the results are compared, a slightly different calculation frame, temporary boundaries, types of works taken into account ... There are different indices: the light of the wedge has not converged on one Thomson, there is an alternative to it Scopus. Their data in some parameters differ almost double.

So, of course, there are different options, more or less honest, and even quite appropriate and official. And there is a sort of "phenomenon of eastern science." It is necessary to talk about him very delicately, because he almost said so - and it turned out that which did not mean at all.

The fact is that Chinese science is growing with great speed. According to the number of publications, China is now next for the United States. This is the second scientific power of the world. As with any rapid growth, it does not do without linings. The scientific journals of the world are littered with a huge number of Chinese works, including many of them - not of the best quality. At the same time, experts said that several years ago the situation was worse, already in a few years the “Chinese articles” became a little more sane, and after time, presumably, the situation will be corrected. But for now, it happens.

Recently, the text appeared on the Web ( www2.madan.org.ua/news/index.php3?bn=MAidan_osvita&key = 1249409309 & Action = View ), which is versed in the example of several scientific journals in mathematical sciences (Applied Mathematics and Computation, Communications in Nonlinear Science and Numerical Simulation, Computer Methods in Applic Mechanics and Engineering, Computers & Mathematics with Applications, Chaos, Solitons & Fractals, International Journal of Computer Mathematics). The authors from different countries of the Middle East and from China are lying into these magazines with a shaft. As far as one can understand, in these countries scientists are paid depending on their “indicators” - publications in magazines with decent impact. In these magazines there is almost no review, in each article there is a very abundant citation by the authors of themselves and their friends. Everyone is satisfied - the authors go for publications, the magazines are increasing citation, the Chinese administration and other countries are proud of scientific growth.

As already mentioned, formal indicators have their own weaknesses. A good article is written longer, sometimes the article begins to quote only after a few years. And indices take into account quoting only “instant”, over two previous years. Innovators are quoted with delay, and the ordinary mainstream is immediately. So the “right” authors lose in indicators, and hacks rower - and knock down the price of the index.

That is, all these international grades and indices can be ruined, and this is already being done. So what remains?

Here we must recall the accompanying circumstances. The science of China is the second largest in the world, with monstrously rapid growth, and Russia with some difficulty, but still overtakes Brazil in the second ten countries. In China, the situation with a large shaft of poor -quality articles is slowly (very slowly), in Russia - gradually becomes worse. Flint with indexes does not go unnoticed - they are not blind around. You can do a lot. You can change the methods of calculations and outpatient not only self -digitalization, but also “circles of quotes”. You can exclude the lists of indexed magazines for the smashed publications. And over time, if one service does not do this (say, Thomson) will make another (for example, Scopus).

You can be upset and worried about the falsification of international indicators of the rating of scientific publications. It is possible, on the contrary, to rejoice that these stories are not covered with darkness and there is the hope that the situation will improve. But actually it is completely indifferent. You can come up with a few "your" pocket indices. Maybe this is not even very bad. But the main thing is not to forget what is really important.

Khoju Nasreddin was somehow asked: when a Muslim has to participate in a funeral procession, where, according to the instructions of Islam, is it best to be-ahead of the funeral stretcher or behind? Khoja answered: it is completely indifferent - ahead or behind, if only not on the stretcher themselves.