Will G Hopkins Auckland University of Technology Auckland NZ

1
Writing Pre Data Scientific WritingGrant
Applications Research Proposals Ethics
Applications
Research Skills Workshop
Writing the summary, common errors, focus,
flow, data. Grants/Proposals content (aim,
significance, review, methods, budget). Ethics
language, risk, subject inducements, deception,
feedback.

• Will G HopkinsAuckland University of
  TechnologyAuckland NZ
• Reviewed/enhanced by Steve Olivier University of
  Northumbria, UK

Minor updates July 2008
2
Scientific Writing

• Your main aim is to make it easy for the reader
  to understand your ideas.
• Search a bibliographic database for what's known
  on the topic(s) in the document.
• PubMed is best for health.
• Google Scholar and SportDiscus are best sport and
  exercise.
• Be obsessional about following any instructions.
  
• Impress reviewers with your ability to obey
  sensible rules.
• Justify any rule breaking.
• Be thorough, fussy, consistent.
• Supervisors, editors and reviewers get angry
  about lack of attention to detail.

3

• The Summary or Abstract is the most important
  part of the document.
• Write it first.
• Rewrite it while or after you write the rest of
  the document.
• Include as much information as possible in the
  allowed space.
• Devote an appropriate amount of material to each
  section.
• Include no new material.
• Avoid abbreviations. Dont make up any.
• Never state the results will be discussed or
  similar.
• Include no references, figures, or tables.
• Avoid these common grammatical errors
• Ambiguous antecedent This means
• Misplaced modifier When injured, we tested the
  athletes.
• Non-human agent This review discusses
• Double doubtful The results may suggest

4

• Avoid also...
• punctuation errors, especially it's and plural's
• too many parenthetical statements
• parentheses (or brackets) within parentheses
  (())
• spelling errors
• abbreviations
• technical terms (jargon)
• hyperbole, such as very, extremely, optimize
• awkward expressions
• passive voice, for example It has been
  identified that...
• wasted words and clichés, such as the results
  showed that...
• wrong words, especially affect and effect
• subject or participant use athlete, runner
• repetition of words, phrases, ideas
• non-sentences
• one-sentence paragraphs

5

• Use of numbers
• Follow the required style. Otherwise, use these
  defaults
• one, two, three nine, 10, 11
• but 3 m, not three meters.
• The test lasted 30 s it was a 30-s test.
• 1500 m, not 1500m
• but 35, not 35
• but body fat was 17 BM (percent of body mass),
  not 17.
• Show data with the appropriate number of digits.
• Standard deviations and the form of confidence
  limits need only two digits or sometimes only
  one.
• The mean then has the same number of decimal
  places 178.4 7.3 or 178 7 1530 170
  (not 1534 173).
• Percents need only two digits 13, 1.3, 0.13.
• Exact P values 0.83, 0.20 0.09, 0.05, 0.01,
  0.007, 0.0003.
• Never Plt0.05 or Pgt0.05.

6

• Aim for focus
• Check that all material is in its most
  appropriate section.
• Have one main idea per paragraph.
• Keep to the topic.
• Don't generalize too far beyond the data.
• Aim for flow
• List headings, subheadings, and topics within
  each section before you begin.
• Avoid unlinked ideas (non-sequiturs).
• Make a logical sequence of ideas within and
  between paragraphs.
• Put technical terms towards the end of sentences?

7

• Reference most assertions.
• Qualify unreferenced assertions with apparently,
  may, I/we believe that
• Get a critical colleague to give you honest
  feedback on a draft.
• Put the final document aside for at least a day
  then read it right through again before you send
  it off.

8

• Mini-test correct the errors...
• The subjects (n34, age 23.45 5yrs, weight 72
  6.2 kg) were recruited from local clubs.
• This suggests the rate of change in heart rate
  response may be important.
• The size of the responses depend on the size of
  the change in heart rate and also the LBNP.
• This study looked at the effect changes of
  pressure had on heart rate.
• The results showed that there was no correlation
  between performance and skinfold thickness
  (r0.234, pgt0.05).
• CONCLUSION Based on this data, the athlete's
  performance was not effected by the supplement.

9
Grant Applications and Research Proposals

• Your institutional research office may be
  helpful.
• Aim/Purpose
• Describe what you hope to achieve.
• Usually the effect of X on Y or the relationship
  between X and Y.
• Do not "determine whether there is an effect",
  because there is always an effect.
• Significance/Benefits
• Explain the originality and the uses of the work.
• State that it hasn't been done before anywhere
  and/or on athletes/people in your country.
• But make sure the novelty of the work is
  sensible.
• Mention potential academic advances.
• Emphasize unique local expertise, resources, or
  opportunities.

10

• Review of Literature
• Keep it focused.
• Cite papers by likely reviewers.
• Methods
• Describe and justify the design. It must be the
  best possible under the circumstances.
• Show a design figure or time line.
• Justify sample size using precision of estimation
  (confidence limits) and sample sizes in
  comparable published studies.
• If the granting body is conservative, show also a
  traditional sample-size estimate, even if it is
  unrealistically large.
• Be specific about what measure will determine
  what outcome.
• Describe or reference techniques and assays.

11

• Demonstrate that you understand the practical
  issues.
• Keep the size of the project manageable.
• Measure a few variables well, rather than many
  poorly.
• Include an account of any pilot work.
• Budget
• Be pessimistic about costs.
• Account for every item.
• Check on the availability and cost of existing
  resources.
• Check on the delivery time of purchases.
• Curriculum Vitae
• Rewrite it to enhance its relevance to the
  proposal.

12
Ethics Applications

• Follow the instructions in your application form.
• Use plain language for the lay members of the
  ethical committee.
• There is no risk-free procedure. Get that across
  or you will be sued when things go wrong.
• Increase subject compliance by providing adequate
  compensation for their time and/or travel.
• If you use deception to avoid placebo effects,
  make sure it is of a kind that the subjects will
  accept when they find out about it at the end of
  the study.
• Try to arrange for all subjects to benefit.
• Commit to high-quality feedback/debriefing for
  subjects.

13
This presentation is available from
Download it from Sportscience 7 (2003).