WHAT DO AUTHORS CARE ABOUT

1
WHAT DO AUTHORS CARE ABOUT?

• What over 50,000 STM Authors Tell Us Each Year

2
Understanding What Authors Say They Want

• Anthropology of scholarly behaviour
• Recording what they say
• Measuring what they do
• Principle of triangulation

3
What Do We Know Already?

• User behaviour research
• Information model for the journal

4
History of User Studies

• 1950s and 1960s
• Merton, Price and Garfield
• 1970s
• Garvey and Griffith, NSF Office of Science
  Communication, Don King, Woolgar and Latour
• 1990s
• Coles, TULIP, SuperJournal, Tenopir and King

5
Robert Merton

• Inventor of the focus group
• Mertons social norms of scientific conduct
• Universalism new work is assessed by universal
  impersonal criteria
• Communality scientific knowledge should be
  common property
• Disinterestedness prime concern is the
  advancement of knowledge
• Organized scepticism knowledge should be
  continually subjected to critical scrutiny
• Reflects stated values rather than actual
  behaviour what they do is not what they say.
• R Merton Sociology of Science, U Chicago P, 1973

6
William Garvey and Belver Griffith

• American Psychological Association research
  surveys into author and reader behaviour
• Early finding about reading
• Survey data suggested journals readings low
• Actually a mistake, failure to scale results from
  the sample to the whole scholarly universe
• Unfortunately contributed to wide-spread library
  myth about low use of journals
• Garvey Griffith Science Communication Amer.
  Psychologist 26(4).14, 1963

7
Time Scales After Garvey
Part of literature
Original material incorporated into texts and
references 12 years
Paper reviewed by annual review volume or
journal 5 years
Oral Report at National Annual Meeting 18-24 month
s
Paper cited in other articles 6 years
Journal Publication 2-3 years
Report to medium-sized restricted audience
Work completed 15 months
Preliminary Oral Report 6 months
Work starts
8
Rhetorical Processes

• Publication is not just communication
• Articles are written to persuade audiences that
• a singular observation made by one observer is
  generally true for all observers at all times
• The research reported is an enactment of the
  idealised scientific method
• Networks of articles collectively construct the
  paradigm pro tempore for the collective
  scientific world view in a discipline
• A G Gross Rhetoric of Science Harvard UP, 1996

9
Rhetorical Status of Research Information
Private Co-workers Invisible
college Speciality Discipline Public
OBSERVATION
Create Discuss revisit Criticism Formal publi
c evaluation Formal confirmation Acceptance
integration
Informal
research
1st draft
COMMUNICATION
Draft for comment
Seminar/workshop/conference
Draft mss
Pre-print
Science journalism
CRITICAL EVALUATION
Peer reviewed paper in a journal
Review paper
ACCEPTANCE AS FACT
reference work
prizes
Formal
monograph
textbook
history
10
Woolgar and Latour

• Anthropological approach to the study of the
  science system
• Steve Woolgar (now at the Oxford Internet
  Institute) and Bruno Latour spent time as
  observers in science laboratories studying the
  behaviour and culture of practising scientists
• First example of an ethnographic approach
• Woolgar Latour Laboratory Life, Princeton UP,
  1979

11
NSF Funded Studies

• Office of Science Communication
• Studies on the alternatives to paper and the way
  the paper system behaves
• Main studies conducted by King Research
• King, McDonald, Roderer Scientific Journals in
  the US, Hutchinson, 1981
• Precursor to Tenopir Kings recent book

12
Methodologies for Studying Behaviour

• Study the Users
• Authors, Editors, Referees, Readers, Librarians
• Opinions and behaviour
• Numbers and groupings
• Study the Outcomes
• Papers, journals, publishers, libraries
• Number, growth and organizational structure
• Ulrichs, ISI, websites etc.

13
Opinions and Behaviour

• Opinions traditional market research
• Questionnaires and Focus Groups
• Research to establish language
• Open not closed questions
• Sample selection
• Channel biases
• Moderators ideally independent
• Behaviour
• Move (inferred use of journal issue)
• Cite (inferred value)
• Download (inferring reading)
• Link (inferred importance)

14
Key Research Studies

• Opinion-based
• Coles 1993
• Elsevier Editorial Strategy Survey 1993-6
• Tenopir King 1995-2003
• Behaviour
• Use various library shelving studies
• Citation analyses Garfield 1960s
• Download behaviour 1993-4 TULIP, 1995-6
  SuperJournal, 2001/2 Nicholas and City University
  Studies
• Outcomes
• De Solla Price, Garfield, and many others

15
Coles 1993 Motivation to Publish
16
Coles 1993 Choice of Journal
quality
speed
collection
habit
17
Elsevier Research 1999 2003

• How do authors choose a journal to publish in?
• They already know the subject coverage of their
  research paper and its quality and approach
• They select the set of most appropriate journals
  in terms of subject coverage
• They match the general quality of their paper
  (best, good, ok) to a class of journals (top,
  middling, run-of-the-mill) with the same subject
  and approach
• From that class they select a specific journal
  based upon experience

18
Elsevier Research 1999 2003

• Most Important Factors
• Reputation
• Refereeing quality
• Refereeing speed
• Impact factor
• Production speed
• Editor/Editorial Board
• Physical quality
• Publisher services

Key Factors Which Category?
Marginal Factors Which Journal?
Journal Hierarchy within a Discipline
Impact Factor Reputation Editorial
Standard Publication speed Access to
Audience International Coverage Self Evaluation
AI Coverage Society Link
Track Record Quality/Colour Illustrations Service
Elements, e.g. author instructions, quality of
proofs, reprints, etc Experience as Referee
J
J
?
A
J
?
J
?
J
J
B
J
J
?
J
J
C
J
19
Constructing a Journal Information Model

• What researchers want as an author
• What researchers want as a reader
• How does a journal deliver this?
• How does the entity responsible for the journal
  do this?
• What are the consequences? Can the model account
  for or predict publishing behaviour?

20
Information Functions of the Journal

• Classical journal functions
• Registration
• Certification
• Dissemination
• Archiving

21
What do researchers want as authors?

• REGISTRATION to register a discovery as theirs
  and made by them on a certain date
• to assert ownership and achieve priority being
  first
• CERTIFICATION To get their research (and by
  implication, themselves) quality stamped by
  publication in a journal of known quality
• to establish a reputation, and get reward being
  in the best journal
• DISSEMINATION To let their peers know what they
  have done
• to attract recognition and collaboration being
  read by all your peers
• ARCHIVE To leave a permanent record of their
  research
• renown, immortality a secure place in the
  literature

22
What do researchers want as readers?

• Reassurance as to its status and quality
• prestige and authority ? CERTIFICATION
• Material that is appropriate to their research
  interest
• specialisation and relevance ? DISSEMINATION
• Tools that allow the material to be located and
  browsed
• browsing and indexing ? NAVIGATION
• Availability of sources over time
• persistence and continuity ? ARCHIVE

23
Behavioural/Functional Model

• Needs
• READERS
• constant citation
• authority
• specialisation
• continuity
• navigation

• Functions
• JOURNAL
• registration
• certification
• dissemination
• archive
• navigation

• Needs
• AUTHORS
• ownership
• reputation
• recognition/audience
• renown

• Provided by the publishing entity through
• third party authority (rhetorical independence)
• brand identity management
• long-term management of continuity
• technology

24
Effect of Nature of Content on the Model
humanities
sciences
25
The Effect of Subject Area on the Model
Subject variation
Small to Medium Scale Experimental
Theoretical Large Scale Experimental
Peer review as methodological and quality filter
Theoretical paper, review re- doestheorem or
proof
Small fields where quality of researchers work
is known to peers
HEP Theoretical Physics Maths
Most quantitative disciplines
Very strong
Very weak
Certification function
26
Effect of Coauthorship Levels on the Model
Pre-print or self-archiving culture?
Unimportant
Registration
Certification
Traditional journal culture
Crucial
1 Level of Co-authorship 100s
27
Author Studies 2003 Results

• Major ongoing study at Elsevier through the
  Author Feedback Programme
• Continuous monitoring of author perceptions via
  questionnaire survey covering all 1200 primary
  Elsevier titles (225,000 sent per year, 79,000
  returned, 35 response rate) in science,
  medicine, technology, social science
• Authors are asked to rate performance of the
  Elsevier title they have just published with
  against their previous journal publishing
  experience
• This allows us to gather comparative data on
  authors irrespective of where they publish

28
Questionnaires An Example
Please indicate whether you agree or disagree
with the following statements concerning the
above journal.
29
Respondents Status and Roles

• Almost all authors are tenured/professional
  researchers
• Hardly any are (graduate) students
• Almost all authors are engaged in RD as their
  main role

30
Respondents Organization

• Majority of authors work in the university
  sector

31
Respondents Age Profile

• Most authors are between 26 and 65 years old,
  with 60 between 36 55
• Age 26 represents first post-doctoral job
• Outside of US most authors retire from 60 65

32
Respondents Productivity

• 74 of authors have published 1.2 10
  papers/yr
• The mode have published about 5/yr

33
Respondents Refereeing Activity

• 83 of authors acted as a referee in the last
  year
• Nearly a third refereed more than 4 papers in
  this period

34
Respondents Editorial Board Activity

• Majority of authors do not serve as editorial
  board members

35
Respondents Priority Ranking
2 1 6 5 7 7 4 3

• Quality
• Speed
• Editor
• Services

36
Network Maturity
37
Network Maturity Rank Order
38
Conclusions

• Most authors are professional researchers in a
  university environment, publishing about 5 papers
  per year. As authors they are very
  journal-focused.
• Most authors act as referees at least once (a
  minority several times) a year and are not
  editorial board members
• They choose to publish from a set of journals
  selected first on specialisation and coverage and
  then subdivided by quality and utility. The
  actual journal chosen depends upon personal
  experience. These choices are intimately
  connected with brand identity issues of journals
  NOT publishers
• They care passionately about the quality and
  speed of the journals they use but not to the
  exclusion of all other factors
• Results are broadly similar across all subjects
  but adoption and comfort of use of IT still
  varies widely