Downloads and Citatations_ Henk Moed_Gali Halevi
-
Upload
gali-halevi -
Category
Science
-
view
111 -
download
0
description
Transcript of Downloads and Citatations_ Henk Moed_Gali Halevi
ANALYSIS OF USAGE AND CITATIONS PATTERNS – SOME OBSERVATIONS
Henk Moed & Gali Halevi
STI 2014 | Leiden
DATA
• Journal Level Data:
• The first set of data contained downloads data for all journals covered in ScienceDirect™ (appr. 1,800) Elsevier full text database.
• Citation data to these journals was extracted from Scopus™
• Both download and citation data relate to the time period 2004-2010, and were aggregated by year and by journal.
• Document Level Data:
• Download and citation counts on a per document basis were collected for all documents published in 62 ScienceDirect™ journals between 2008 and 2012 covering all domains of science and scholarship.
• Downloads and citations counts on document level are up to September 2013.
• User sessions: Data related to downloads from ScienceDirect made by two European academic institutions during 2002-2003.
• User institutions: Data on downloads from ScienceDirect of documents published during January 2009-May 2013 made from selected institutions during the same time period.
• User countries: Data on downloads from ScienceDirect of documents published during January 2009-May 2013 made from selected countries (China and UK) during the same time period.
Terms
• Usage : any type of activity relating to the use made of electronic publication archives in the broadest sense, and recorded in the archive’s electronic log files.
• It includes activities such as downloading in pdf, viewing in html format, browsing through abstracts, and also saving, sharing or annotating documents in reference managers.
• Download is accessing the full text, either in PDF or in HTML format, of a document indexed in Science Direct.
USER LEVEL OBSERVATIONS
Download counts from three account countries (ScienceDirect)
Seasonal variations: USA and UK tend to have holidays in August; China in February
Effect of National Research Assessment Excercises
Download counts from Italy (ScienceDirect)
Monthly downloads for 3 user institutions: large variations over time [Moed & Halevi, JASIST, 2014a]
ARTICLE LEVEL OBSERVATIONS
Downloads vs. citations (article from J. of Wind Energy and Industrial Aerodynamics)
Citations (red curve, low numbers)
The effect of citations upon downloads
Corrected proof online 04-03-2008
Corrected paginated proof online 22-08-2008Downloads
11
Number of Downloads as a function of the articles’ age CLIN MED
LIFE SCI
MATH
APPL SCI
HUMAN
The number of downloads per document type as a function of the documents’ age
REVIEW
EDITORIAL
Ratio of accumulated downloads and citations of documents as a function of their age
FULL LENGHT
EDITORIAL
REVIEW
DISCIPLINARY LEVEL OBSERVATIONS
Ratio of accumulated downloads versus citations for full length articles in 62 journals
A&H
SOCIAL SCI.
MATH
Materials Chem Vs. Management
ScienceDirectjournals only
17
Usage vs. citations per main field
Authors < Readers
Authors = Readers
?PSYCHOL
ScienceDirect journals only
18
Good coverage
Moderate coverage
CONCLUSIONS
10 important factors differentiating between downloads and citations [Moed &
Halevi, JASIST, 2014a]
1 Usage leak: Not all downloads may be recorded.
2 Citation leak: Not all citations may be recorded.
3 Downloading the full text of a document does not mean that it is read.
4 The user (reader) and the author (citer) population may not coincide.
5 Distribution # downloads less skewed than that of # cites, and depends upon the type of document differently
6 Downloads and citations show different obsolescence functions.
7 Downloads and citations measure distinct concepts.
8 Downloads and citations may influence one another in multiple ways.
9 Download counts are more sensitive to manipulation.
10 Citations are public, usage is private.
http://emdp.elsevier.com
The program supports individuals and research groups working on scientific evaluation metrics by offering data and financial sponsorship.
The aim of the program is to encourage research into Bibliometrics, Altmetrics, individual or institution level metrics and others
THANK YOU