Dissertation Information for Wanda Maria Paranhos NAME: - Wanda Maria Paranhos
DEGREE:
- Ph.D.
DISCIPLINE:
- Library and Information Science
SCHOOL:
- Case Western Reserve University (USA) (1981)
ADVISORS: - William M. Shaw Jr.
COMMITTEE MEMBERS: - William Goffman - [Indecipherable1] - Miranda Pao
MPACT Status: Complete - Except Indecipherables
Title: Application of an Entropy Measure for Journal Evaluation and its Comparison to Other Quantitative Measures
Abstract: This dissertation addresses the problem of journal evaluation by means of quantitative measures. A set of 856 journals indexed in 1978 by Index Medicus and by the Citing Journal Package of Journal Citation Reports constituted the data base. The first objective of this study was to identify titles in this -population of journals which seemed to contribute most to the communication process observable in it. Journals were studied as elements of a communication network and their relationships were expressed in terms of reliance. Reliance is the fraction of the total citations given by a journal which is given to a cited journal. Journals were clustered on the basis of interreliance, which is the extent to which a pair of journals rely on each other at a given threshold. Journal were clustered according to the single-link clustering technique. The Brillouin information was used to identify synthetic journals, which are said to contribute most to the communication structure of the network. There were 184 titles identified as synthetic journals at thresholds .004 through .028, in increments of .004 . A comparison was then made between the results obtained with this entropy measure and the results produced by four other quantitative measures. The other measures were productivity, citations received, impact factor, and influence measure. A high positive correlation was found between citations received and influence measure, suggesting that they are redundant. The other results were too weak to allow anyone measure to predict the relative standing of a journal according to another measure. A comparison between top ranking journals according to each of the' five quantitative measures studied and six core lists of biomedical journals was also made. Results of chi square tests indicated that top ranking titles identified by each quantitative measure were more likely to be included in core lists than lower ranked titles. Citations received presented consistently better results. There was a variation in the identity of the top ranked journals according to the five measures studied. A comparison was also made between the journal ranking by the five quantitative measures studied and a peer review ranking, by subject category, of some of the titles. Agreement between the peer ranking and the ranking produced by at least one of the quantitative measures was found for most categories. Overall impact factor seemed to produce statistically significant results more consistently, while productivity seemed to produce the least significant results. The findings suggest that the property of the journal literature which is associated with the Brillouin information measure is also worth consideration in a journal selection process They also suggest that this process still profits from the consideration of the different assessments made by most of the measures studied. Recommendations for further study are given.
| |
MPACT Scores for Wanda Maria Paranhos A = 0
C = 0
A+C = 0
T = 0
G = 0
W = 0
TD = 0
TA = 0
calculated 2008-01-31 06:01:20
Advisors and Advisees Graph
generating graph, please reload |