ISSN:
1572-9486
Source:
Springer Online Journal Archives 1860-2000
Topics:
Physics
Notes:
Conclusions The foregoing analysis assumes the whole of physics journal literature to follow a Bradford distribution. However, figure 1, as well as other published Bradford plots of physics literature, show similar “Groos droop” deviations from linearity at the upper end [13, 14]. The common explanation for this deviation is incompleteness of the bibliographic data [5], but this of course cannot be proven without compiling a truly complete world bibliography on the subject. If the droop is real, i.e. not the consequence of bibliographic incompleteness, then the estimates of journals and papers are exaggerated. On the other hand, another assumption is that the first 110 ranked journals in this study are indeed the top producing physics journals, and that all source items published in these journals are actually physics papers. The extent to which this assumption does not hold means a decrease in the estimates, thereby opposing the first possible error. The fact that the actual number of papers published by the journal of first rank exceeds B, the theoretical number derived from analysis, suggests an underestimate of total journals and papers. This supports the assumption that any deviation from the Bradford distribution is minimal and that the Groos droop is a consequence of bibliographic incompleteness. It cannot be assumed that the highest quality literature is concentrated in the higher Bradford ranks. There may be a relationship between quality and scatter, as earlier work has suggested [9], but this is a matter for further investigation. It is clear that frequency-rank distributions such as Bradford's Law have more potential for scientific information management than has been so far exploited.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF01599718
Permalink