Transcript Publication Title

Impact Factors:
how to get one,
how to improve it,
how to understand it.
ISAJE Annual Meeting 2nd
September 2006
Elaine Stott
(Journal Publishing Director,
Blackwell Publishing)
Impact Factor
• Eugene Garfield created The Science Citation Index (SCI) in
1961 as a quarterly printed index, - 600 journals.
• Aim to look at relationships between researchers – who was
citing who?
• Journal Impact Factor was used for journal selection.
• Never intended to be used as it is today – librarians, grants,
RAE in UK, authors
2
The Impact Factor
The average number of citations the average article
receives per annum in the 2 years after publication
e.g. the 2005 impact factor will be the total citations in
2005 to articles published in 2003 and 2004 / no. articles
published in 2003 and 2004
Citations
Article published in Jan 2003 will have longer to contribute cites to 2005 IF
than an article published in Dec 2004
0
2
4
6
8
10
12
14
16
18
Time after publication (Years)
3
20
Document Type
Letters: Rapidly climbs to citation peak, large proportion of cites in 2 year window – but low absolute number
Full Papers: Later citation peak, smaller proportion in 2 year window, but greater longevity
Reviews: Latest citation peak of three document types, and greatest longevity. Proportion of cites within
impact window is low - but high absolute number of cites
4
Variation in Impact Factors (Mabe & Amin, 2000)
Significant variation in mean impact
according to subject area
Significant effect of multiple authorship
Can be field specific.
Top journal in one field may have a lower impact
than bottom journal in another field.
Comparisons should only be made within the same subject areas
5
Impact Factor Fluctuation: Size of Journal
Correlation between impact factor fluctuation and size of journal
A journal of 200 papers per year might expect up to ± 15 % fluctuation
For an impact of 1.5 this would mean a range of 1.27 -1.72 implies no significant change
6
The Immediacy Index
A measure of skewness of the citation curve
i.e. how quickly are the articles being cited?
No. of citations in the current year / No. articles published in that year
Citations
Some evidence that the II appears to be correlated with IF, i.e. a
change in II is followed by a change in IF – useful early warning
signal.
0
2
4
6
8
10
12
14
16
18
Time after publication (Years)
7
20
The Cited Half-Life
A measure of rate of decline of the citation curve
i.e. article longevity
The number of years that the total citations (i.e. citations to
articles from all years, rather than only the previous two years)
takes to decline to 50% of its value
Citations
No evidence that the cited half-life is correlated with IF
50% of
citations
0
2
4
50% of
citations
6
8
10
12
14
16
18
Time after publication (Years)
8
20
Getting an Impact Factor
• Today covers 8,500 of a total of 20-30,000 journals (c.
•
•
•
35% of all journals).
ISI aim to include most internationally influential journals
in each subject area
2,000 journals evaluated every year -both new +already
included jnls
150-200 new added each year
9
Getting an Impact Factor
• ISI aim to include most internationally influential journals
•
•
in each subject area
2,000 journals evaluated every year -both new +already
included jnls
150-200 new added each year
Details on ISI website:
http://scientific.thomson.com/mjl/selection/#jsc
10
Selection criteria for inclusion in ISI database
• Basic Journal Publishing Standards
•
•
•
11
a. Timeliness of Publication?
b. International Editorial Conventions?
c. English Language Bibliographic
Information?
d. Peer Review?
Editorial coverage – does the journal add
anything new, how does it compare to other
journals waiting to be added?
International diversity ( take International + best
national Journals)?
Number of citations
Analysis of citations
• Related journals
• Top cited institutions
• Top cited authors
• Where are top cited papers being published
• Which subjects are getting most citations
12
Addiction journals
Addiction journals are listed in Journal Citation Reports:
Science
Substance abuse – ( 8 journals) IF range 0.9 -3.6
Psychiatry (84 Journals) IF range 0.05 -12.6
Pharmacology and Pharmacy (193 Journals) IF range 0.1 19.83
Toxicology (75 Journals) IF range 0.2-19.3
Psychology (60 Journals) IF range 0.1 - 9.7
Social Science
Substance abuse – (18 Journals) IF range 0.5 - 3.6
Psychology, general – (111 Journal) IF range 0.02 - 9.7
Psychology, clinical (53 Journals) IF range 0.05 -5.0
Social Issues (31 Journals) IF range 0.7 -1.6
13
2005 Impact factors – Science -substance
abuse
1
ADDICTION
3.696
2
DRUG ALCOHOL DEPEN
2.969
3
ALCOHOL CLIN EXP RES
2.636
4
ALCOHOL ALCOHOLISM
2.04
5
ADDICT BIOL
1.846
6
ALCOHOL
1.743
7
J STUD ALCOHOL
1.654
8
SUBST USE MISUSE
0.983
14
2005 Impact factors – Social Science substance abuse
1
ADDICTION
3.696
2
PSYCHOL ADDICT BEHAV
2.098
3
J SUBST ABUSE TREAT
2.077
4
EUR ADDICT RES
1.667
5
J STUD ALCOHOL
1.654
6
DRUG ALCOHOL REV
1.618
7
ADDICT BEHAV
1.581
8
AM J ADDICTION
9
ALCOHOL RES HEALTH
1.179
10
AM J DRUG ALCOHOL AB
1.094
11
SUBST USE MISUSE
0.983
12
J ADDICT DIS
0.894
13
J DRUG ISSUES
0.812
14
J PSYCHOACTIVE DRUGS
0.752
15
J DRUG EDUC
0.741
16
DRUG-EDUC PREV POLIC
0.619
17
ADDICT RES THEORY
0.594
J CHILD ADOLES SUBST
0.579
18
15
1.47
Analysis of citations
• Related journals
• Top cited institutions
• Top cited authors
• Where are top cited papers being published
• Which subjects are getting most citations
16
Papers 2002-2006 in impact factor journals
Total articles Total citations
98
256
Finland
17
Addiction
30
63
Alcohol &
Alcoholism
29
78
ACER
19
65
Substance
Use &
Misuse
7
1
DAD
4
27
Alcohol
3
11
Addiction Biol
3
4
J. Sub. Abuse
3
7