Top five in physics
Novel citation index highlights hot topics.
Are you working on the hottest topic in your field? Many scientists may think so, but it has been a tough assertion to prove — until now, that is. A German physicist has devised a way of answering the 'Hot or not?' question for his discipline. If it stands up to scrutiny, it could be used to rate topics across the sciences. In physics, the results show that hotness — measured by a parameter known as m — correlates well with the promise of future wealth... and that promise is greatest in nanotechnology.
Find the physics top five here. (You'll need a password)
Disagree with the results? Tell us why.
Comments
I agree that these topics are pretty hot, but are they "physics"? I would more readily classify most of them as chemistry/materials science.
Posted by: Ivan Tubert-Brohman | May 18, 2006 06:24 PM
These "hot topics" are all narrowly defined things, as opposed to novel theories, or new physical phenomena. Has physics become stamp collecting too? I don't think so.
Posted by: Preston MacDougall | May 19, 2006 01:26 PM
What about HTc superconductors?
Posted by: Marco Zangrando | May 19, 2006 03:34 PM
The topics are more materials Science and Engineering than Physics. Especially nanotubes, and nanowires. The papers are primarily new fabrication methods and not any new physics.
Posted by: Kalpana Katti | May 22, 2006 08:11 PM
The image by S. HEINZE as related to 12.85 Carbon nanotubes may be profound. At once it resembles the ballistic rifling of a gun barrel and the helical mechanics of quantum particles and celestial bodies. With a little imagination one may visual Wilson lines and loops that seem to suggest the possibility that quantum loops can be integrated into helical loops or a type of twistor string theory.
Posted by: Doug | May 23, 2006 05:03 PM
My comments regarding the h-b index are:
First, can one really measure the importance of a topic by how many researchers are actively working on it? Take the field of medicine as an example, is working on cancer research more important than working on cardiovascular diseases just because there are more active researchers in the former research area than the latter or vice versa?
Second, Web of Science indexes a little over 300 Physics journals. What percentage of all Physics journals does this figure constitute? Will the small number of Physics journals indexed in Web of Science have any impact on the results of h-b index and m scores? Studies that examine the breadth of coverage of databases indicate that the single most comprehensive database in the majority of fields account for only one-third of the indexed literature in its respective field (see Meho and Spurgin, 2005). As for citation databases, a study in progress shows that using Elsevier's Scopus, in addition to Web of Science, not only increases the citation counts of individual scholars in Computer Science and Information Science by an average of one-third but also significantly alters their relative rankings (Meho & Yang, in progress). The point here is how can one tell whether or not a topic is "hot" or "important" when the database used to determine this covers only a small portion of world literature on that topic? It should be emphasized here that broadening the sources of data may not alter results in all of research fields, but it certainly does in many of them (Meho & Spurgin, 2005; Moed, 2005).
Third, to accurately measure h-b and m scores for a topic, it must be with certainty that (1) all relevant records on the topic were retrieved from the database--i.e., 100% recall; and/or (2) the percentage of recall across different topics is relatively the same. Both of these tasks are impossible to achieve rendering h-b scores inaccurate at best. H-scores, on the other hand, rely on whether or not an item was cited; only in rare cases, one fails to retrieve all relevant citations for a particular author, work, or journal, among others (Cronin & Meho, 2006).
Fourth, there are millions of documents in Web of Science that do not have abstracts (review articles are a good example and so are most documents from the Arts & Humanities file within the database). One, therefore, cannot say that a topic is more "important" than another unless all or most records in that topic have abstracts in Web of Science or the data source used.
Fifth, another concern regarding the use of abstracts for retrieving relevant documents on a topic is the question whether are all abstracts, even within the same research topic, area, or discipline, are written the same way or use the same structure. Do most or all of these abstracts succeed in mentioning or capturing all the important keywords in a paper? It is quite possible that missing relevant and highly cited records due to poorly written abstracts or lack of abstracts altogether could significantly alter h-b scores.
Sixth, I believe that to argue that a topic is more important or hotter than another, one has to identify all possible topics in a research area or discipline and then compute the h-b scores for all before making any conclusions. The h-b scores in this case will only work using a pre-determined list of topics, and conclusions should be limited to these topics only rather than making generalizations to an entire field.
Seventh, in many fields, it takes three or more years for an article to get published and then indexed in Web of Science. Therefore, one cannot say that the h-b index accurately predicts hot topics. What was "hot" three years ago may not be "hot" today or remain hot in the future. In the same token, topics that may be hot today will take them years before Web of Science and h-b will be able to identify them as "hot" or "important." Taking the h-index as an example, there is no doubt that it is a very hot topic today and has been so for almost a year. As of today, however, there are only 10 or so articles published on the topic and are indexed in Web of Science. Of these 10 articles, only one is cited so far. If we apply the h-b and/or m formula on the topic of h-index, it will not show any significance.
In short, the idea of the h-b index is great, but there are so many factors that weaken its reliability, applicability, and generalizability.
Cited References:
Cronin, Blaise, & Meho, Lokman I. (2006). Using the H-index to Rank Influential Information Scientists. Journal of the American Society for Information Science and Technology, 57(9).
Hirsch, J.E. (2005). An Index to Quantify an Individual's Scientific Research Output. Retrieved September 29, 2005, from http://xxx.arxiv.org/abs/physics/0508025.
Meho, Lokman I., & Kristina M. Spurgin. 2005). Ranking the Research Productivity of LIS Faculty and Schools: An Evaluation of Data Sources and Research Methods. Journal of the American Society for Information Science and Technology, 56 (12), 1314-1331.
Meho, Lokman I., & Yang, Kiduk. (in progress). The Beginning of a New Era in Citation Analysis, Bibliometrics, and Scholarly Communication: The Impact of Scopus and Google Scholar. Intended for publication in the Journal of the American Society for Information Science and Technology.
Moed, Henk F. (2005). Citation Analysis in Research Evaluation. Dordrecht: Springer.
Lokman I. Meho, Ph.D.
Assistant Professor
School of Library and Information Science
Indiana University
1320 East 10th Street, LI 011
Bloomington, IN 47405-3907
Tel: (812) 856-2323
Fax: (812) 855-6166
E-mail: meho@indiana.edu
http://www.slis.indiana.edu/faculty/meho/
Posted by: Lokman I. Meho | May 28, 2006 04:59 AM
This is an edited version of a comment I posted on the same topic on May 28, 2006.
I read with interest your article about a new way of measuring the popularity of different scientific fields. Devised by Michael Banks from the Max Planck Institute for Solid-State Physics, it says that a topic with an "hb index" of 10 will have at least 10 published papers, each of which has been cited at least 10 times. By dividing this value by the number of years that papers on that topic have appeared, Banks obtains a separate parameter, m, which indicates the topic's relative importance. This method makes carbon nanotubes the most popular field, with an m of 12.85.
I see a number of flaws with his method. First, although it may determine the most popular subjects, it certainly does not measure the importance of a topic.
Second, Banks uses data from Thomson's Web of Science database, which indexes a limited number of physics journals -- roughly 300. I am not sure what fraction of all physics journals this accounts for, but I have found that, for most fields, the single most comprehensive database accounts for only one-third of the indexed literature of the field [see, for example, Lokman I. Meho and Kristina M. Spurgin, "Ranking the Research Productivity of LIS Faculty and Schools: An Evaluation of Data Sources and Research Methods," Journal of the American Society for Information Science and Technology, vol. 56, no. 12 (2005), pp. 1314-1331)].
Moreover, my studies have shown that using Elsevier's Scopus, in addition to Web of Science, increases the citation counts of individual scholars in Business, Computer Science, and Information Science by an average of one-third and also significantly alters their relative rankings. Therefore, by limiting the analysis to Web of Science, the assessment of a topic's "heat" or importance is based on only a small and most likely skewed portion of world literature on the topic.
Banks's approach also overlooks the problem that millions of documents in the Web of Science database -- review papers, for example -- do not have abstracts at all. Moreover, can he be sure that all abstracts, even within the same research topic field, include all of the important keywords in the paper? It is quite possible that his method ignores relevant and highly cited papers that contain poorly written abstracts.
In addition, to argue that one topic is more important than another, one should identify all possible topics in a particular discipline and compute the hb scores for each before making any conclusions. The scores in this case will only work using a pre-determined list of topics, and conclusions should be limited to these topics only rather than making generalizations to the entire field.
Finally, it should be noted that it takes a long time (in many cases over a year) for an article to get published and then indexed in Web of Science. What was "hot" a year or more ago may not be hot today or in the future. Similarly, topics that are hot today may take years before they appear in Web of Science and/or get cited.
In short, the idea of the hb index is great, but there are many factors that weaken its reliability, applicability and generalizability.
Lokman Meho, Ph.D.
School of Library and Information Science, Indiana University, US
meho@indiana.edu
Posted by: Lokman I. Meho | July 6, 2006 04:41 AM
The hot picks for the physics field are definitely not physics related, most of those were chemistry or materials picks. Furthermore, in physics, we consider anything that is fullerene or SWNT related as being on its way out, since that field has been around since the late 80's. The newer stuff in physics is like Bose-Einstein condensates, laser physics, nanophotonics, and anything that has to do with high T superconducting materials. Also biophysics is up and coming, as is AFM work. Work with gold-silver is considered hot in physics.
Posted by: Betty Catalina Rostro | July 25, 2006 05:22 AM