Writing peer-reviewed publications Professor Robin Room

1
Writing peer-reviewed publicationsProfessor
Robin Room
2
1. Have an idea

• A conceptual piece? A commentary?
• Only some journals are interested
• Probably easier for senior scholars to get
  published
• May not count as peer reviewed
• A review of the literature?
• increased demands that it be systematic
• Journals like systematic reviews they get cited
  a lot
• A meta-analysis?
• Are there enough more-or-less comparable studies?
• A qualitative study whats the question?
• a conceptual orientation needed not just a
  slice of life
• need to specify methods, etc.
• A quantitative analysis whats the question?
• appropriate analytical methods
• bite-size how much can fit in one paper?

3
1a. From report for a government department to
peer-reviewed publication

• The report as framed for a government department,
  for instance, is unlikely to make a good journal
  article
• Switch your thinking
• from whats happening in this population/place?
• to what is of general interest to the field?
• Whether methods, patterns, relationships,
  testing hypotheses
• Situate your thinking/writing in the literature
• Pick a journal you are aiming the publication at,
  and read its instructions to authors

4
2. Introduction / background

• Give an early indication of your topic and
  orientation
• Discuss the relevant literature
• Many of us get hung up on the literature review
  dont obsess
• You are checking what has been done relative to
  your idea
• In the paper, you need to summarise previous
  findings enough to situate your paper
• Dont put endless references without any
  indication of why
• Indicate gaps in the literature as well as what
  is known
• Ideally, the literature review should point to
  the need for your analysis to fill a gap
• Clearly state what the paper is going to do

5
3. (for an empirical paper) Material and methods

• Describe the material of the study
• Compress, but be very concrete
• Sample frame and design, size, completion rate
• If the frame is a clinical or other
  social-handling selection process, something on
  how people get there
• Figure out ways of giving the actual wording of
  questions or categories (e.g. in tables)
• Refer to technical report etc. (if available), or
  previous papers on the same data, for further
  details
• Statistical methods
• If you are doing something pretty standard, this
  can be very brief

6
4. Findings

• A. Quantitative
• You have done a lot of runs, now the problem is
  how to compress as much data as possible into
  four tables or so
• One table to set the stage sample composition,
  etc. but try to have it more than just this
• Sometimes, the stage-setting is better done in
  the text
• With logistic regressions, etc., show multiple
  models
• Once you have the tables, writing up what is in
  them is a snap
• But make sure there is text about all the
  analyses implied by the tables

7
The point of multiple models

• Allows you to sort predictors by conceptual
  (potential causal) status (different disciplines
  have different terms for this)
• Prior to everything/explanatory (but are
  demographics always prior?)
• Intermediating/interpreting
• Specifying
• Comparing results with and without a tranche of
  predictors
• e.g., attitudes ? problems, vs.
• attitudes ? consumption ? problems
• An example predicting entry to alcohol treatment
  in Stockholm ...

8
(No Transcript)
9
Be inventive about summarising
Beer or spirits inherently more harmful? -- It
depends Significant differences in consequences
of drinking, comparing beverages in surveys in 19
countries. gt means more consequences than ?
per 1000 grams of alcohol. ? for those whose
drinking is gt2/3 of the type.
No. significant/no. of comparisons ? 15/55 ? 40/82
Beer gt wine SE, DK, CA CR, BL, PE, SE, DK, CZ, IM, CA, AU, NZ
Wine gt beer NC, KA  
Beer gt spirits UG, BR CZ, CA
Spirits gt beer CZ, CA BL, IN, KA
Wine gt spirits NC, KA  
Spirits gt wine SE, DK, CZ, CA CR, BL, KA, DK, CA, AU, NZ
AR (Argentina), AU (Australia), BL (Belize), BR
(Brazil), CA (Canada), CR (Costa Rica), CZ
(Czech Republic), DK (Denmark), IM (Isle of Man),
IN (India), KA (Kazakhstan), NC (Nicaragua), NG
(Nigeria), NZ (New Zealand), PE (Peru), SE
(Sweden), SR (Sri Lanka), UG (Uganda), UR
(Uruguay)
10
4. Findings (continued)

• B. Qualitative
• Sort the material into themes, which should add
  together into a story
• Enough quotes to give verismo
• Keep quotes relatively short
• Dont treat qualitative data on a convenience
  sample as if it were quantitative
• OK to say most, a few, about half, but
  dont give
• Its more convincing if you acknowledge not
  everything fits together
• give an example or two of counter-instances

11
5. Discussion and conclusion

• One section or two?
• usually two, but then give them different
  functions
• Make sense of the findings
• but not by parading your prejudices
• Fit them into the previous findings in the
  literature
• Refer back to the material in the Introduction.
• Mention limitations of the study
• What are the next steps?
• Summarise towards the end

12
6. Around the edges

• Orient your style to the journal you are aiming
  for!
• Length of piece, number and format of tables
• Reference style
• Abstract style
• Authorship
• If to be multiply authored, have some agreement
  beforehand
• Keep within guidelines for qualifying for
  authorship
• Mostly, junior researchers have more to gain than
  senior ones from first authorship

13
Around the edges (continued)

• Reviewing a topic
• Good publication material from an editors
  perspective high citation rate
• Suit the approach/method to the topic and its
  literature
• Conceptual/analytic? Systematic? Meta-analysis?
• Commentaries, Editorials, Book reviews
• Often commissioned who does the editor know?
• Make yourself and your interests known
• Propose before writing?

14
Around the edges (contd)

• Choosing a journal
• Impact factors and their discontents
• In principle, the average number of citations per
  article in a two-year period after publication
• Jostling for position how to improve the factor
• Alcohol/drug/gambling journals have relatively
  low rankings
• But dont drive yourself crazy, dont overreach
• What audience would you like to reach? Still a
  relevant question
• Whether in PubMed, ISI, etc. indexes
• But less important now because of Scholar.google
• Problems for qualitative, historical, policy
  analyses
• Online open-access journals -- the wave of the
  future?

15
Source Babor, Stenius Savva, Publishing
Addiction Science, 2004.
16

• The pecking order in 2007 (not a great deal of
  change since)

Some current impact factors Lancet
38.28 JAMA 23.20 AmJPubH 3.93 MJAust
2.89 ANZJPH 1.20
17
The tyranny of the two-year window
Rank Impact Factor 2009 Impact 2005-09 Impact 1981-2009
1 Addiction(3.84) Addiction(8.03) British Journal of Addiction(25.56)
2 Journal of Substance Abuse Treatment(2.90) Journal of Studies on Alcohol(7.01) Journal of Studies on Alcohol(24.70)
3 Psychology of Addictive Behaviors(2.68) Alcoholism Clinical and Experimental Research(6.69) Journal of Substance Abuse(21.21)
4 International Journal of Drug Policy(2.54) Drug and Alcohol Dependence(6.67) Addiction(21.15)
5 Addictive Behaviors(2.25) Addiction Biology(6.31) Research Advances in Alcohol and Drug Problems(20.90)
6 Journal of Studies on Alcohol and Drugs(2.07) Psychology of Addictive Behaviors(5.38) Alcoholism Clinical and Experimental Research(20.57)
7 Drug and Alcohol Review(1.65) Alcohol(5.35) Drug and Alcohol Dependence(17.21)
8 American Journal on Addictions(1.53) Journal of Substance Abuse and Treatment(5.10) Addictive Behaviors(15.44)
9 American Journal of Drug and Alcohol Abuse(1.34) Alcohol and Alcoholism(4.54) Alcohol(15.21)
10 Journal of Gambling Studies(1.28) Addictive Behaviors(4.22) Psychology of Addictive Behaviors(15.10)
The table compares the citation impact of
journals in substance abuse as measured over
three different time spans. The left-hand column
ranks journals based on their 2009 "impact
factor," as enumerated in the current edition
ofJournal Citation Reports. This is calculated
by taking the number of all current citations to
source items published in a journal over the
previous two years and dividing by the number of
articles published in the journal during the same
period--in other words, a ratio between citations
and recent citable items published. The rankings
in the next two columns show impact over longer
time spans, based on figures from Journal
Performance Indicators. Here, total citations to
a journal's published papers are divided by the
total number of papers that the journal
published, producing a citations-per-paper impact
score over a five-year period (middle column) and
a 29-year period (right-hand column).
SOURCE Journal Citation Reports and Journal
Performance Indicators. http//www.sciencewatch.co
m/dr/sci/11/apr10-11_2/
18
Addiction journal or disciplinary journal?
Source Babor, Stenius Savva, Publishing
Addiction Science, 2004.
19
Reference sources for our field

• Available on the web
• Babor, Stenius Savva, Publishing Addiction
  Science, 2004
• http//www.parint.org/isajewebsite/isajebook/isaje
  webbook.htm
• Miller, Strang Miller, Addiction Research
  Methods, 2010
• http//au.wiley.com/WileyCDA/WileyTitle/productCd-
  1405176636.html