Transcript Professor Gavin Reynolds

Publishing your research
Professor Gavin Reynolds
Honorary Professor: Biomedical Research Centre, Sheffield Hallam University, UK
Professor Emeritus: Queen’s University Belfast, UK
Visiting Professor: SouthEast University, Nanjing; University of Malaya, Kuala
Lumpur; Naresuan University, Thailand
Past-President: British Association for Psychopharmacology
Publishing your research
•
•
•
•
•
•
What you need for a great paper
Understanding how biomedical research is done
Select the right journal
How to write a paper
How to write to the editor
How to respond to reviewer’s comments
So you want a great paper!
• To have a great paper you need to have:
• Great results!
– Which need to be novel, clear and meaningful.
– But to get great results you need to have:
• A great project!
– Which needs to be well designed, adequately powered.
– But to have a great project, first you need:
• A great idea!
How biomedical research is done
•
•
•
•
•
Identify the problem
Understand the background
Propose the hypothesis
Design the study
Do the work
• Collect the sample, make the measurements
• Analyse the data
• Write and publish the report
• Identify the next problem...
How biomedical research is done
– an example
The problem:
• Some drugs used to treat schizophrenia make people
gain weight.
• This can cause increases in diabetes and heart disease
• People have different responses – some get very fat,
some have no weight increase, even when they
receive the same drug.
How biomedical research is done
– an example
Understanding the background:
• Antipsychotic drugs used to treat schizophrenia
have effects on the brain.
• Our brain controls how much we eat.
• There are receptors in the brain affecting appetite
that these drugs may block.
• Differences between people in weight gain may be
due to genetic differences.
How biomedical research is done
– an example
Developing the hypothesis:
• There are genetic differences between people in
weight changes after drug treatment for
schizophrenia.
• These differences can be found in genes controlling
eating.
“Variability in the gene for the 5-HT2C receptor is
associated with weight gain during antipsychotic
drug treatment.”
How biomedical research is done
– an example
Design and do the study:
• Find patients with schizophrenia receiving drugs for
the first time.
• Measure changes in weight over 10 weeks.
• Investigate genetic differences between patients in
one gene important in controlling appetite.
How biomedical research is done
– an example
Analyse the results:
• Genetic differences
between patients were
related to the amount of
weight gain.
• Patients with one genetic
variant had much less
weight gain.
Period of drug treatment
How biomedical research is done
– an example
Write and publish the report:
How to select the right journal
• What is the “right” journal?
–
–
–
–
–
–
The right readership – specialist or general?
The highest impact factor? Included in Pubmed?
Previously published related papers?
Rapid publication?
Appropriate for style/format of report?
Sympathetic to Asian manuscripts?
How to write a paper
• Having great results is not enough. You need to:
– Understand how your work relates to work from other
groups
– But don’t forget that others in your research group may
be doing related work – understand that too!
• Write for the reader
– What is interesting for you may not be for the reader
– Decide what your main message is, and focus on that
– Ask a colleague to read it through and criticise
• Be aware of other related papers
– Always think about how your work adds to, or improves
on, other research.
How to write a paper
• Read and follow the journal instructions
– Look at the layout of published papers. You can use other
papers as a “template” for your report
– But do not “copy and paste” from other papers! Editors
have software that can detect this! And referees hate it
when they see their words repeated by someone else!
– Be tidy! Use page break to separate sections, figures etc.
– Tables and figures in separate files or at the end of text file
– Always define abbreviations the first time you use them:
“Brain-derived neurotrophic factor (BDNF) is… We measured BDNF…”
– Format references correctly
– Correct format for abstract and text headings
How to write a paper
• Introduction
– Most journals prefer the introduction to be just a brief
summary of the background leading up to your work
– Go from the general to the particular:
Start with a general statement of the context of your work:
“Depression is common in patients with Parkinson’s disease, who
frequently respond poorly to antidepressant drugs”
– End with the particular focus of your research:
“Thus we aimed to determine whether this genetic polymorphism is
related to the response to depression of patients with PD”
– Include a statement of your hypothesis or research question
How to write a paper
• Methods
– For human samples, list inclusion and exclusion criteria,
indicate “written informed consent” and ethical approval
– For in vivo animal studies, always indicate ethical standards.
– Experimental methods can be brief if they replicate
published methods: “…following the method of Yang et al (2011)…”
or “… based on the method of Yang et al (2011). In brief, this involved…”
How to write a paper
• Results
– Include all results you wish to report, but do not interpret
or comment on them.
– Refer to results in tables/figures, but do not repeat the
data in the text. You can report it in another way (e.g. %).
– Age, sex etc. of a sample are results, not for the methods
section, unless predetermined in the sample collection
• Statistical results
– Be accurate in describing the data
– Not “…was decreased” but “…was significantly decreased”
– Not “…was not decreased” but ”…was not significantly
decreased”
1. Statistically highly significant difference; “is significantly much less”.
2. No difference between the mean values – “no difference” is only correct here.
3. Difference is less, but does not reach statistical significance. Not: “not different”
How to write a paper
• Discussion
– Good to start with a simple summary of the main
findings, but don’t just repeat the results.
– Relate your findings to your hypothesis.
– Discuss your findings in relation to other work in the
field, but don’t repeat the introduction.
– Be aware of the limitations and discuss them, before the
reviewer asks you to do that!
– Draw conclusions
How to write a paper
• Common language errors
– Affect and effect: affect is usually a verb; effect is usually a
noun, meaning consequence or result of some action
– Information, never informations
– Evidence, never evidences
– Research, never researches
– Data is plural; singular is “data point” or “datum” (rarely
used).
• Think about tense – present or past
– “Wang et al (2007) found that cognition is impaired after
methamphetamine administration…”
– past perfect to describe what was done
– present if the finding is generally true.
How to write to the editor
• Be brief – it is enough just to say:
– “We wish to submit our paper entitled… for publication
in …”
• Remember that the editor will read the abstract
to understand the content of the paper
– If you need to explain your work in the letter to the
editor, then you have not written the abstract clearly!
How to write to the editor -II
• But you can emphasise important aspects if it is
particularly notable:
– “We believe the finding that … has important
implications for future diagnostic testing and treatment”
• Remember that the editor is interested in
increasing the impact factor for his journal
– “As the results may be seen as controversial, they should
attract substantial interest and citations”
How to respond to the reviewers
• You respond to the reviewers’ comments
but you write to the editor.
– So you can point out to the editor if the reviewer is wrong,
or has missed something, or is being inflexible over a
disputed point
• Make your response clear and brief - include each of
the reviewer’s points in your response and address it:
– The authors have not included the units in table 3
– We have now corrected this error
– They have failed to cite the report that this polymorphism
is associated with depression (Smith et al)
– The reviewer has missed our mention on p12 of several
reports failing to replicate the finding of Smith et al.
How to respond to the reviewers - II
• If you believe a reviewer to be biased, or incorrect,
don’t be afraid to say so to the editor:
– “While the reviewer appears to reject our findings on the
basis of the work of Smith et al., this is a view that is not
common to more recent studies we have cited (e.g. Jones
et al)…”
• I usually finish by saying something like:
– “We hope you agree that we have appropriately addressed
all the reviewers’ concerns and believe the manuscript is
now suitable for publication.”
Please ask me questions!