|
Guidance on reviewing a paper
An evaluation
of a scientific article is not simply a summary of the article but should
consider the soundness of the methodology and any particular weaknesses
of the study. Scientific articles usually have five distinguishable
sections (abstract, introduction, methods, results and discussion).
The evaluation should be based on your personal constructive comments
towards the merits of the article. This can be easily judged by evaluating
each of these sections individually. The following are examples of the
sorts of questions you should ask.
INTRODUCTION
 Is the study original?
Are aims and objectives
clearly stated?
Are they reinforced
by suitable citations of previous work and how useful were they (i.e.
is the literature review extensive)?
METHODS
How was the study conducted?
Subject characteristics
- age, sex, race, health status, experimental setting.
Criteria for admission
or rejection from study.
Exposure -source,
route, other possible confounding exposures.
Methods (subjective
or objective) and equipment used to collect data. (if appropriate were
they single or double-blinded).
If appropriate were
the used methods ethically evaluated?
Were experiments
adequately controlled?
Was the study design
sensible?
Does it clearly
explain the used statistical methods?
Was systematic bias
avoided or minimised?
Was the study large
enough or long enough to make the results credible?
Was the technical
analysis used suitable for this kind of exposure?
Were the data analysed
according to the original study protocol?
RESULTS
Were the results presented in a logical fashion?
Have these been
clearly expressed (e.g. in the form of figures or tables)?
Were the data limited
by the sensitivity of the recording devices?
Were there an adequate
number of observations made?
DISCUSSION
Are the conclusions
clearly stated?
Are they positive
or negative (or unrelated) in relation to the aims?
Are they justified
by the experiments or undue extrapolations?
Does the study point
to further lines of research?
Does the study discuss
the internal consistency of the data?
Is there a discussion
of the relevance of the results in relation to other research?
How could it have
gone wrong? (bias, confounding?)
Do the results fill
the aim of the study?
Are there any clear
tables that display the results?
SUMMARY
(OR ABSTRACT)
Does it encompass
the main points?
Are statements or
conclusions reported in the summary qualified by statements of uncertainty
(e.g. authors sometimes make simplified conclusions in the summary,
whereas the discussion reveals that several possible alternative explanations
have been considered)?
OVERALL
JUDGEMENT
This is necessarily tempered by the weight assigned to deficiencies
in the various sections discussed above. If you are critical of the
experimental design, you should balance this by assessing the practicality
of the available alternatives.
Nachaat
Tahmaz
DEOM, University of Aberdeen
The
views and opinions expressed in the text are solely
of the review author
Back
to Journals Review Page
|
