23 januari 2007
Web of Science reviewed by Jacsó
My favourite database reviewer still is Péter Jacsó. His digital reference shelf is a column to keep an eye on. This month he reviews Web of Science. Since the introduction of the Citation reports it was about time.
The review is very thorough and I really would love to learn some of his tricks which he uses in his database reviews as well. It was a pleasure to digest it all. One of those things that I never realized was the fact that abstracts were only included systematically from 1991 onwards. I knew they were more often lacking from the older material but this sharp demarcation line in the change of policy was not known by me. It should have been good if Jacsó had pointed out the changes in naming policy as well. Once it was 15 characters, later 18 and nowadays first names are indexed as well. All this makes comprehensive searches for long standing researchers sometimes difficult. Well, challenging at least. Especially when you're dealing with double names. "F.C.T. Penning de Vries" is one of my favourites. He can be found as DEVRIES P in the cited references. ISI has promised author disambiguation, but this has not reached the science commons yet. And Penning de Vries was not a bad scientist after all. On Author disambiguation Scopus does a better job.
The new citation reports in WoS are swell indeed. The h-index (not Hirsch index so Hirsch told us) implementation is good indeed, and very useful to apply to all kind of search results. Journals amongst others, as Jacsó did for some of the LIS journals. I also hope to see these results included in the Journal citation reports next year.
But after all this praise some grumps as well. It is a rather old one. Sloppy indexing by ISI. Jacsó came with the example of author names. Also issues, volumes and page numbers go wrong quite often. Although Moed (2005, p.175) has established that this is only in the order of about 7-8%.
We use WoS quite often within the library setting itself for collection development. We look at our authors and examine their reference lists. That runs quickly into the tens of thousands references, and we would love to see some uniformity in the journal names of those references. This is clearly illustrated when you do a search for a journal in the cited reference search. With the cited reference search you should use the abbreviated journal name (as opposed to the full search where full journal names are requested, but I am not complaining). Take for instance the AUSTRALIAN JOURNAL OF AGRICULTURAL RESEARCH, the official abbreviation is AUST J AGR RES. But don't dare to think that on entering you have found all instances of this journal. Oh no!. You have found only 33,276 citations for this abbreviation. But you missed AUST J AGR RESEARCH (1), AUST J AGRIC RES (1158), AUST J AGRIC RS (23), AUST J AGRIS RES (1) AUSTR J AGR RES (11417) AUSTR J AGR RESEARCH (1) AUSTR J AGR RESER (6) AUSTR J AGR RS (2) AUSTR J AGRI RES (24) AUSTR J AGRIC RES (127) AUSTR J AGRIC RESEAR (1) AUSTR J AGRICULT RES (6) AUSTR J AGRICULTURAL (251?) AUS J AGR RES (35) AUS J AGRIC RES (3) AUS J AG RES (19) AUST J AG RES (90) AUST J AGIC RES (3) and there are probably a few other variations I missed. My point is however, if you analyze a large number of references you inevitably end up with a lots of variations of journal names. This doesn’t only apply for this particular instance. We try to monitor the usage of about 8000 journals. With sloppy data as in the above example it becomes a real tour de force, which I would love to see better facilitated on the side of ISI. Obvious corrections in the data should therefore be done on the primary data, rather than in the software which takes place to some extent.
But to end this post on a positive note, I quote Jacsó's conclusions in full
References
Jacsó, P. (2007). "Web of Science." Peter's digital reference shelf Retrieved Jan 2007 from http://www.gale.com/reference/peter/200701/wos.htm.
Moed, H. F. (2005). Citation analysis in research evaluation. Dordrecht, Springer. 346p.
The review is very thorough and I really would love to learn some of his tricks which he uses in his database reviews as well. It was a pleasure to digest it all. One of those things that I never realized was the fact that abstracts were only included systematically from 1991 onwards. I knew they were more often lacking from the older material but this sharp demarcation line in the change of policy was not known by me. It should have been good if Jacsó had pointed out the changes in naming policy as well. Once it was 15 characters, later 18 and nowadays first names are indexed as well. All this makes comprehensive searches for long standing researchers sometimes difficult. Well, challenging at least. Especially when you're dealing with double names. "F.C.T. Penning de Vries" is one of my favourites. He can be found as DEVRIES P in the cited references. ISI has promised author disambiguation, but this has not reached the science commons yet. And Penning de Vries was not a bad scientist after all. On Author disambiguation Scopus does a better job.
The new citation reports in WoS are swell indeed. The h-index (not Hirsch index so Hirsch told us) implementation is good indeed, and very useful to apply to all kind of search results. Journals amongst others, as Jacsó did for some of the LIS journals. I also hope to see these results included in the Journal citation reports next year.
But after all this praise some grumps as well. It is a rather old one. Sloppy indexing by ISI. Jacsó came with the example of author names. Also issues, volumes and page numbers go wrong quite often. Although Moed (2005, p.175) has established that this is only in the order of about 7-8%.
We use WoS quite often within the library setting itself for collection development. We look at our authors and examine their reference lists. That runs quickly into the tens of thousands references, and we would love to see some uniformity in the journal names of those references. This is clearly illustrated when you do a search for a journal in the cited reference search. With the cited reference search you should use the abbreviated journal name (as opposed to the full search where full journal names are requested, but I am not complaining). Take for instance the AUSTRALIAN JOURNAL OF AGRICULTURAL RESEARCH, the official abbreviation is AUST J AGR RES. But don't dare to think that on entering you have found all instances of this journal. Oh no!. You have found only 33,276 citations for this abbreviation. But you missed AUST J AGR RESEARCH (1), AUST J AGRIC RES (1158), AUST J AGRIC RS (23), AUST J AGRIS RES (1) AUSTR J AGR RES (11417) AUSTR J AGR RESEARCH (1) AUSTR J AGR RESER (6) AUSTR J AGR RS (2) AUSTR J AGRI RES (24) AUSTR J AGRIC RES (127) AUSTR J AGRIC RESEAR (1) AUSTR J AGRICULT RES (6) AUSTR J AGRICULTURAL (251?) AUS J AGR RES (35) AUS J AGRIC RES (3) AUS J AG RES (19) AUST J AG RES (90) AUST J AGIC RES (3) and there are probably a few other variations I missed. My point is however, if you analyze a large number of references you inevitably end up with a lots of variations of journal names. This doesn’t only apply for this particular instance. We try to monitor the usage of about 8000 journals. With sloppy data as in the above example it becomes a real tour de force, which I would love to see better facilitated on the side of ISI. Obvious corrections in the data should therefore be done on the primary data, rather than in the software which takes place to some extent.
But to end this post on a positive note, I quote Jacsó's conclusions in full
"ISI has kept adding new content and software features through regular updates. The latest services clustering of results set by several criteria, the instant calculation and superbly informative and compact visualization of new citation measures, such as the sum of times a paper was cited (including and excluding self-citation, the average citations per item, the Hirsch-index, the almost instant display of charts for the distribution of articles and citations per year by authors, journals, organizations or topic, the exporting of these details into a spreadsheet format, or downloading to a free Web version of EndNote, represent more than a series of evolutionary steps. It is a breakthrough for those interested in citation analysis, but did not have the resources to calculate key citation performance measures, or did not have the software to format them to the whims of the publishers' manuscript guidelines."
References
Jacsó, P. (2007). "Web of Science." Peter's digital reference shelf Retrieved Jan 2007 from http://www.gale.com/reference/peter/200701/wos.htm.
Moed, H. F. (2005). Citation analysis in research evaluation. Dordrecht, Springer. 346p.
Labels: English
Comments:
Een reactie plaatsen
Links to this post:
<< Home
Links to this post:
<< Home
