Overview

1
The Challenge of Scientific Writing (i)
Content
2
The Challenge of Scientific Writing (ii)
Content
Writing
3
The Challenge of Scientific Writing (iii)
Concise Coherent Clear
Complex Unambiguous
Nonreferential Unobtrusive
4
The Challenge of Scientific Writing (iv)
Concise Coherent Clear
Complex Unambiguous
Nonreferential Unintrusive
Content
5
The Challenge of Scientific Writing (v)
To stay within the required format and still say
something interesting.
6

Je n'ai fait celle-ci plus longue que parce que
je n'ai pas eu le loisir de la faire plus
courte. I have written a long letter because I do
not have time to write a short one. - Blaise
Pascal, "Lettres provinciales", Letter 16,
December 4, 1656 http//www.ac-nice.fr/philo/text
es/Pascal-Provinciales.htm
7
Uniform Requirements

• Follow them.
• A misformatted submission is a red flag.
• Rejection for nonadherence is unusual
• If the journal deviates, you deviate

8
Usual Format
Original contributions

• Size (text) 2,500 to 5,000 words
• Size (abstract) 150 to 250 words
• Illustrations 3 to 6 Tables, Figures, Graphics

9
Overall Structure
Original contributions
I
Introduction (Why did you do it?)
M
Methods (How did you do it?)
R
Results (What did you learn?)
D
Discussion (Who cares?)
10
Introduction (i)
Original contributions

• Key point SHORT
• State the issue (3-5 sentences)
• State the commonly held belief (4-6 sentences)
• State what you will do (2-4 sentences)

11
Original contributions
Introduction (ii)
Example The rationale underlying contact
investigation for tuberculosis (TB) is that
certain clinical presentations pose a risk of
infection to patient contacts through inhalation
of airborne droplet nuclei containing
Mycobacterium tuberculosis.1,2 This rationale
has generated a set of priorities for contact
tracing justified by the frequency of
transmission of M. tuberculosis to close contacts
of infectious patients. For example, contacts of
patients with non-cavitary, or smear-negative
pulmonary TB would have a lower priority than
contacts of patients with sputum AFB
smear-positive pulmonary disease. Contacts with
evidence of latent TB infection (LTBI) can reduce
their risk of progression to active disease by
60-70 upon completion of a 6 to 9 month course
of isoniazid.3 In recent years, primarily as a
result of the HIV and STD epidemics, interest has
grown in the contribution that social network
analysis can make to understanding disease
transmission.3-5 Several examples of its
application in the field of TB control suggest
that a network-informed approach holds promise
for improving both the understanding of
transmission dynamics and the effectiveness with
which secondary TB cases and contacts with LTBI
may be discovered through the contact tracing
process.6,7 The underlying hypothesis for a
network-informed approach to contact
investigations is that in areas of ongoing
transmission, this strategy will provide access
to a wider group of persons involved in a
transmission milieu. An infectious persons
contacts may have TB or LTBI because of direct
contact with him or her, or by having acquired it
elsewhere. Thus, a network investigation, it
might be predicted, would be able to uncover a
group that is epidemiologically and biologically
interconnected (that is, have social, sexual, or
drug-using connections and have M. tuberculosis
isolates with matching DNA fingerprint patterns)
as well as persons with epidemiologic
connections, but unique DNA fingerprint
patterns. In this study, we explore the use of
network-informed approaches, coupled with RFLP
typing, in the investigation of a protracted TB
outbreak in Wichita, Kansas in a group of 19
young persons (25-35 years of age) diagnosed with
TB over a seven year period. The outbreak was
first identified among women who worked as exotic
dancers, and we investigated the role of social,
sexual, and drug-using network relationships in
understanding TB transmission
12
Original contributions
Introduction (iii)
Example (The Issue) The rationale underlying
contact investigation for tuberculosis (TB) is
that certain clinical presentations pose a risk
of infection to patient contacts through
inhalation of airborne droplet nuclei containing
Mycobacterium tuberculosis.1,2 This rationale
has generated a set of priorities for contact
tracing justified by the frequency of
transmission of M. tuberculosis to close contacts
of infectious patients. For example, contacts of
patients with non-cavitary, or smear-negative
pulmonary TB would have a lower priority than
contacts of patients with sputum AFB
smear-positive pulmonary disease. Contacts with
evidence of latent TB infection (LTBI) can reduce
their risk of progression to active disease by
60-70 upon completion of a 6 to 9 month course
of isoniazid.3 (Common Beliefs) In recent years,
primarily as a result of the HIV and STD
epidemics, interest has grown in the contribution
that social network analysis can make to
understanding disease transmission.3-5 Several
examples of its application in the field of TB
control suggest that a network-informed approach
holds promise for improving both the
understanding of transmission dynamics and the
effectiveness with which secondary TB cases and
contacts with LTBI may be discovered through the
contact tracing process.6,7 The underlying
hypothesis for a network-informed approach to
contact investigations is that in areas of
ongoing transmission, this strategy will provide
access to a wider group of persons involved in a
transmission milieu. An infectious persons
contacts may have TB or LTBI because of direct
contact with him or her, or by having acquired it
elsewhere. Thus, a network investigation, it
might be predicted, would be able to uncover a
group that is epidemiologically and biologically
interconnected (that is, have social, sexual, or
drug-using connections and have M. tuberculosis
isolates with matching DNA fingerprint patterns)
as well as persons with epidemiologic
connections, but unique DNA fingerprint
patterns. (What you will do) In this study, we
explore the use of network-informed approaches,
coupled with RFLP typing, in the investigation of
a protracted TB outbreak in Wichita, Kansas in a
group of 19 young persons (25-35 years of age)
diagnosed with TB over a seven year period. The
outbreak was first identified among women who
worked as exotic dancers, and we investigated the
role of social, sexual, and drug-using network
relationships in understanding TB transmission
13
Methods (i)
Original contributions

• Key point
• Stands alone

14
Original contributions
Methods (ii)
Overall structure
15
Original contributions
Methods (iii)
Provide the context 1-2 paragraphs give
general background for study describe setting
in which study was done (do NOT provide
justification)
16
Original contributions
Methods (iv)
Describe what you did 7-10 paragraphs sample
selection and ascertainment instruments
used biologic measurements (procedures) analyt
ic approaches (data management) (human
subjects considerations)
17
Original contributions
Methods (v)
Reference standard methods 1-2
paragraphs cite analytic methods cite
statistical and computer tools cite standard
data bases
18
Original contributions
Methods (vi)
Acknowledge Human Subjects issues One
paragraph mention use of informed
consent mention IRB approval
19
Original contributions
Methods (vii)

• Some donts
• Do not include methods considered but abandoned
  (dont recreate the thinking)
• Do not justify the methodological choices
• Do not discuss their implications
• Do not offer results or opinion

20
Original contributions
Results (i)
Key point Dont tell them everything you know.
21
Original contributions
Results (ii)

• Overall structure
• Results should be parallel with Methods (roughly)
• Proceed from simple to more complex
• General description
• Univariable (unadjusted) comparisons
• Multivariable (adjusted) comparisons
• More complex models and statistical approaches
• Mention only other approaches that were
  confirmatory or noncontributory

22
Original contributions
Results (iii)

• Illustrative material (i)
• avoid tables if you can say it in the text
• avoid text tables whenever possible
• avoid graphs if you can use tables
• use legends to graphs instead of footnotes
• Avoid duplication in tables and text

depending on overall length and specifications
of the journal
23
Original contributions
Results (iii)

• Illustrative material (ii)
• Do not use a table, graph, or figure that is not
  referred to (cited) in the text
• Cite tabular content, not the table itself
• Wrong The comparative heights of men and women
  are shown in Table 3.
• Right In general, men are taller than women
  (Table 3).

24
Original contributions
Results (iii)

• Illustrative material (iii)
• Avoid appendices if possible
• Do not include data tables, questionnaires
• Include numerical examples, proofs, etc.
• Avoid footnotes
• Incorporate material into text

25
Original contributions
Results (iv)

• Illustrative material (iv)
• Avoid graphs and figures with a low
• contentink ratio

26
Original contributions
Results (v)

• General coherence
• Do not use a technique in the Results that is not
  mentioned in the Methods
• Do not mention something in the Results that you
  do not consider in the Discussion

27
Original contributions
Discussion (i)
Key point Dont go beyond the data.
28
Original contributions
Discussion (i)

• Overall structure (i)
• First Paragraph
• Use the first paragraph to restate the basic
  question and the major result(s) that illuminate
  the issue
• Do not use the first paragraph to summarize the
  findings

29
Original contributions
Discussion (i)
Overall structure (ii) Discussion sections
30
Original contributions
Discussion (i)

• Overall structure (iii)
• 3-5 Major points
• Offset each point visually, using either a 2nd
  level header or by beginning a paragraph with an
  italicized phrase.
• To the extent possible, maintain parallelism
  between the Discussion, the Results, and the
  Methods.

31
Original contributions
Discussion (i)
Overall structure (iv) Example Discussion This
is the opening paragraph that restates for the
reader what the major research questions is and
what the importance is of this study. A major
point This is the first paragraph that discusses
a major point. This is the second paragraph that
discusses a major point. A second major
point This is the first paragraph. This is the
second paragraph.
32
Original contributions
Discussion (i)
Overall structure (v) Alternative
example Discussion This is the opening
paragraph that restates for the reader what the
major research questions is and what the
importance is of this study. A major point.
This is the first paragraph that discusses a
major point. It uses an italicized phrase at the
beginning of the sentence to offset the
point. This is the second paragraph that
discusses a major point. A second major point.
This is the first paragraph of second major
point. It uses an italicized phrase at the
beginning of the sentence to offset the
point. This is the second paragraph that
discusses a second major point
33
Original contributions
Discussion (i)

• Overall structure (iv)
• Strengths and Limitations
• Understate. Dont break your arm patting
  yourself on the back.
• Avoid phrases like This is the first study to
  do such-and-such.
• Include real limitations, not strawmen (items
  that sound like limitations but are really
  strengths).
• Avoid phrases like This study has several
  limitations.
• Defend the methods used, as appropriate

34
Original contributions
Discussion (i)

• Overall structure (v)
• Implications of the findings
• Understate. Do not go beyond the data
• Provide biological plausibility for findings, if
  appropriate
• Point to parallels in other lines of inquiry, if
  appropriate

35
Original contributions
Discussion (i)

• Overall structure (vi)
• Future directions
• Try to say something that enhances the readers
  perspective on the issue, based on the presented
  findings.
• Avoid saying More research is needed. (If more
  research is needed, be more specific.)
• Draw conclusions if warranted by the data

36
Original contributions
Discussion (i)

• Overall structure (vii)
• Extraneous Sections
• Do not include a separate section for Conclusions
  (these are integrated into the Discussion)
• Do not include a section called Summary (that is
  what the Abstract is for)

37
Original contributions
Discussion (ii)

• General coherence (i)
• Dont include results that you do not discuss.
• Dont discuss findings that were not presented in
  results. (That is do not introduce new material
  in the discussion.)

38
Original contributions
Discussion (ii)

• General coherence (ii)
• Keep Methods, Results, and Discussion
• roughly parallel
• mutually exclusive
• Dont put results or discussion in Methods
• Dont discuss the findings in the Results
• Dont put methods or results in the Discussion

39
Original contributions
Structured abstract (i)

• Key points
• Write the abstract AFTER you write the paper
• Remember that is probably the only part of the
  paper that most people will read

40
Original contributions
Structured abstract (ii)

• Generally 150-250 words
• Several alternative formats
• Check the uniform guidelines and the specific
  instructions from the journal to which the
  manuscript will be submitted

41
Original contributions
Structured abstract (iii)
Example (i) (from Sexually Transmitted Diseases)
Background Goal Study Design Results
Conclusions
42
Original contributions
Structured abstract (iii)
Example (ii) (from JAMA) Context Objective
Design and Setting Participants Main
outcome Measure Results . Conclusion
43
Original contributions
Structured abstract (iii)
Example (iii) (from IJE, NEJM) Background
Methods Results Conclusions
44
Original contributions
Structured abstract (iii)
Example (iv) (from American Journal of Public
Health) Objectives Methods Results
Conclusions
45
Original contributions
Structured abstract (iii)
Example (v) (from Annals of Epidemiology)
Purpose Methods Results Conclusions
Notes Do not include references in the
abstract Do include numeric results, if applicable
46
Title (i)
Original contributions

• Not too long
• Effect of glycoprotein IIb/IIIa receptor blockage
  with abciximab on clinical and angiographic
  restenosis rate after the placement of coronary
  stents following acute myocardial infaction. J
  Am Coll Cardiol 200035915-921

• Not too short
• Occupational asthma. J Allergy Clin Immuno
  2001108317-328

• Just right
• A prospective study of obesity and risk of
  coronary heart disease in women. NEJM
  1990322882-889

47
Title (ii)
Original contributions

• The colon () approach
• The effects of pre-emptive treatment of
  postherpetic neuralgia with amitriptyline a
  randomized, double-blind, placebo-controlled
  trial. J Pain Symptom Manage 199713327-331

• The question
• What to do with HLA-DO?. Immunogenetics.
  51(10)765-70, 2000

• The statement
• Supporting community initiatives is crucial to
  scaling up orphan support activities in Africa.
  Int Conf on AIDS, Barcelona, Spain, abstract

48
Title (iii)
Original contributions

• The inscrutable reference
• The key and the lamppost. J Clin Oncol
  1999173007-3008

• Terminally cute
• Recidivism redux. Sexually Transm Dis.
  200026350-352

• The neverending story
• Users' guides to the medical literature XIV.
  How to decide on the applicability of clinical
  trial results to your patient. JAMA.
  1998279(7)545-9

49
Title (iv)
Original contributions

• Clever and meaningful
• Telling tails explain the discrepancy in sexual
  partner reports. Nature. 1993365(6445)437-40
• telling tails telling tales
• Chronicle of an epidemic foretold. Millbank
  Quarterly 199371565-574
• (despues de Cronaca di una morte annunciata de
  Gabriel García Márquez)

50
Original contributions
Housekeeping (i)

• The numbering of the tables should agree with
  citations in the text
• Figures quoted in text and tables should agree
• All the references in the reference list should
  be quoted in the text. Conversely, no citation
  in the text should be missing from the reference
  list
• Do not reference material that would not be
  available to the reader
• Personal communications and articles in
  preparation should be cited in the text (keep to
  a minimum)
• Keep citations in press to a minimum

51
Original contributions
Housekeeping (i)

• Spellcheck
• Remove all word processing word processing
  detritus.
• Align the left, but not the right, margin
• Use the journals style sheet for size,
  paragraphs, citations, superscripts, subscripts,
  etc.
• See Uniform Requirements (handout)
• Send the required number of copies

52
Original contributions
Housekeeping (ii)
Please do not use the cover letter as a marketing
device.
53
Original contributions
Housekeeping (iii)
Example (i) Dear Editor We are enclosing the
article Collisions with passenger cars and
moose, Sweden, for your consideration. We
believe this article is the most comprehensive
summary yet attempted of adverse moose-human
interactions. We have conducted an extensive
epidemiologic investigation of each reported
incident, with post-tracking of other moose in
the area to observe further moose behavior and
determine whether or not the general comportment
of moose in the area is conducive to
accidents. We appreciate your prompt
consideration of this manuscript and look forward
to its acceptance and publication. Sincerely (
NOTE This article was published in the Am J
Public Health. 76(4)460-2, 1986) the letter is
fictitious, with apologies to the authors)
54
Original contributions
Housekeeping (iii)
Example (ii) Dear Editor We are enclosing the
article Collisions with passenger cars and
moose, Sweden, for your consideration. This
article demonstrates the importance of human
interactions with wildlife species that are
present in a given environment. The specifics
of human-moose interactions provide some examples
of how the manner in which space is shared with
wildlife can be improved. We appreciate your
consideration of this manuscript and look forward
to learning of your assessment. Sincerely (NOT
E This article was published in the Am J Public
Health. 76(4)460-2, 1986) the letter is
fictitious, with apologies to the authors)
55
The Scope of Scientific Presentation
Other Formats The Birth Announcement The
Laboratory Report The Social Science paper
56
Birth Announcement
The Birth Announcement (i)

• Usual format
• Has all the properties and characteristics of an
  Original Publication
• IMRD format
• About 3,000 words with 4-6 illustrations

57
Birth Announcement
The Birth Announcement (ii)

• Content
• Describes the establishment of a new study, often
  a multi-center undertaking
• Often provides no results other than a
  description of an enrolled population (often less
  than that)
• Occasionally contains material of generic or
  methodological interest

58
Birth Announcement
The Birth Announcement (iii)

• Problems
• A great convenience to authors who get a
  publication and can then cite their methods
• Of relatively low interest to Editors and
  journals, since there is little new or
  interesting
• Little assurance that the actual results of the
  study will be sent to the same journal

59
Laboratory Report
The Laboratory Report (i)

• Usual format
• Has all the properties and characteristics of a
  Brief Report
• IMRD format
• About 1,500 words with 1-3 illustrations
• Usually has a substantial Methods section so
  that the experimental results can be reproduced
• Usually has minimal discussion
• See accompanying materials

60
Laboratory Report
The Laboratory Report (ii)
Content Reports the results of one or a series
of experiments, often resulting in a single
addition to the literature. Generally very
short, and not meant to be read, but to be filed
and consulted
61
Laboratory Report
The Laboratory Report (iii)
Significance A lab report is the type of
material that requires rapid dissemination and
may be the type of article that is most suited
for online publication
62
Social Science Report
The Social Science Manuscript
Usual format Size (text) 4,000-6,000
words Size (abstract) 150-250 (often
unstructured) Illustrations variable, but
usually many
63
Social Science Report
The Social Science Manuscript
Key point A different dialecta different
approach to the presentation of data
64
Social Science Report
The Social Science Manuscript
Comparison with medical manuscripts
65
Social Science Report
The Social Science Manuscript
Content (i)
66
Social Science Report
The Social Science Manuscript

• Content (ii)
• A much longer, more detailed presentation
• An embedded review of the literature and of the
  major issues
• A more speculative discussion (often) of the
  ramifications of the subject matter
• The expectation of a more in-depth consideration

67
Social Science Report
The Social Science Manuscript

• Content (iii)
• Both approaches have merits and drawbacks
• The two approaches set up a potential culture
  clash, most clearly seen when an investigator
  sends a manuscript to the wrong journal.

68
SPUorSalami styling

• Smallest publishable unit
• A single idea, observation, factoid etc. that the
  author thinks he or she can make into a full
  paper
• Salami styling
• Seeing how many papers a single coherent document
  can be divided into

69
Microstructure
The Seven Deadly Sins of Scientific Dialect

• i Passive voice
• Hesitant and unconvincing
• Seems to be concealing something
• Takes more words to say
• Lends itself to twisted sentence constructions
• Example

When the condition of the patient at the time of
discharge was suggestive of an undiagnosed wound
infection, an effort was made to trace the
patient through local nursing homes and family
physicians.
When the patients condition at discharge
suggested an undiagnosed wound infection, we
tried to trace the patients through local nursing
homes and family physicians.
70
Microstructure
The Seven Deadly Sins of Scientific Dialect

• ii Hedging (overqualifying) to death
• Candor, caution, and modesty are virtues
• Dont overdo it
• Keep the caution in the content, not in the
  modifiers
• Example

It appears that this finding might possibly
suggest a potentially new approach to analysis.
It (i)appears that this finding (ii)might
(iii)possibly (iv)suggest a (v)potentially new
approach to analysis.
This finding may suggest a new approach to
analysis.
71
Microstructure
The Seven Deadly Sins of Scientific Dialect

• iii Big-deal words
• Sometimes a big word captures the exact meaning
• It backfires if you use a big word that adds
  nothing
• Instead of being precise, it makes you appear to
  be showing off
• Examples

72
Microstructure
The Seven Deadly Sins of Scientific Dialect

• iv Noun strings (i)
• A string of consecutive nouns, each of which acts
  as a modifier, presumably of the final one
• Acceptable, and even required, in German
• In English, it creates confusion, since it is not
  always clear which are modifiers and which are
  the important nouns
• Examples

• Computer systems analyst
• Health information system retrieval protocol
• Human resource development specialist
• Health promotion material dissemination services
• Human capital investment paradigm

73
Microstructure
The Seven Deadly Sins of Scientific Dialect
iv Noun strings (ii) Examples
Diabetic patient blood pressure reduction may be
a consequence of renal extract depressor agent
application.
Applying depressor agents from renal extracts may
reduce blood pressure in diabetic patients.
The procedure must produce a valid analysis of
the diabetes in pregnancy adverse outcome
prevention program.
The procedure must produce a valid analysis of a
program for preventing adverse outcomes resulting
from diabetes during pregnancy.
74
Microstructure
The Seven Deadly Sins of Scientific Dialect

• v Verbs weak or smothered (i)
• Inactive or linking verbs (to be, to have) are
  weak
• They are smothered by prepositions and
  qualifying phrases
• The verb is effectively turned into a noun
• Examples

There is a need to investigate these association
in additional populations for further elucidation.
These associations should be investigated in
other populations.
The identification and classification of the
various histologic types of lymphomas are vital
steps toward the introduction of new therapies
and the reduction of mortality.
Classifying the histologic types of lymphomas may
lead to new therapies and reduced morality.
75
Microstructure
The Seven Deadly Sins of Scientific Dialect

• v Verbs heaped (ii)
• Strings of verbs, each of which adds only a shade
  of meaning
• Particularly favored in bureaucrat-ese
• Often motivated by a fear of incompleteness
• Examples

We were attempting to construct, develop, and
implement a program to prevent contamination of
the blood supply.
We started a program to prevent contamination of
the blood supply.
76
Microstructure
The Seven Deadly Sins of Scientific Dialect

• vi Tack-ons (i)
• One form of misplaced modifier
• Tacking onto the end of a sentence a phrase that
  begins with
• compared with
• resulting in
• based on
• Examples

Varying prevalences are observed in other groups,
resulting in lower R-squared values
The varied prevalences in other groups result in
lower R-squared values.
77
Microstructure
The Seven Deadly Sins of Scientific Dialect

• vii Redundancy (i)
• Wordiness or repetition
• Can apply to sentences, paragraphs, or whole
  sections
• probably the most important source of
  unnecessary length

78
Low content-to-ink-ratio
79
High content-to-ink ratio
80
New Courses at Emory in Science Writing
JRNL 488-0PF (SHORT COURSE) Science Writing for
Research May 22-26, 2006 (Monday, May 22 through
Friday, May 26)930 a.m.-1230 p.m.
JRNL488-1PF  (SHORT COURSE)  Science
Journalism May 22-26, 2006 (Monday, May 22
through Friday, May 26)900 a.m.-1200 pm.