Measuring similarity of concentration between different informetric distributions: Two new approaches
Article first published online: 18 MAR 2005
Copyright © 2005 Wiley Periodicals, Inc.
Journal of the American Society for Information Science and Technology
Volume 56, Issue 7, pages 704–714, May 2005
How to Cite
Burrell, Q. L. (2005), Measuring similarity of concentration between different informetric distributions: Two new approaches. J. Am. Soc. Inf. Sci., 56: 704–714. doi: 10.1002/asi.20160
- Issue published online: 7 APR 2005
- Article first published online: 18 MAR 2005
- Manuscript Revised: 17 MAY 2004
- Manuscript Accepted: 17 MAY 2004
- Manuscript Received: 19 FEB 2004
From its earliest days, much investigative work in informetrics has been concerned with inequality aspects. Beginning with the well-known Gini coefficient as a measure of the concentration/inequality of productivity within a single data set, in this study we look at the problem of measuring relative inequality of productivity between two data sets. A measure originally proposed by Dagum (1987), analogous to the Gini coefficient, is discussed and developed with both theoretical and empirical illustrations. From this we derive a standardized measure—the relative concentration coefficient—based on the notion of “relative economic affluence” also introduced by Dagum (1987). Finally, a new standardized measure—the co-concentration coefficient, in some ways analogous to the correlation coefficient—is defined. The merits and drawbacks of these two measures are discussed and illustrated. Their value will be most readily appreciated in comparative empirical studies.