Proposing and Communicating the Results of Evaluation Studies: Ethical, Legal, and Regulatory Issues - Evaluation Methods in Biomedical Informatics - page 352

Writing Objectivist Study Reports
The format of objectivist study reporting has evolved over the last century
into a well-known structure represented by the IMRAD acronym. This
format is primarily useful for communicating evaluation study results
to technical/scientific audiences. Such a report includes the following
components:
1. Introduction to the problem, review of relevant literature, and state-
ment of study goals.
2. Methods employed, including details of statistical tests used, ideally
described in enough detail (or by reference to published papers) to allow
another investigator to replicate the study.
3. Results, often summarized in tables or graphs. Some audiences, includ-
ing some professional journals, now ask for full data to be sent with the
article for the purposes of refereeing and public access. With the authors'
agreement, these data can be made available to other interested parties.
4. Analysis or interpretation of the data.
5. Discussion of the results and potential limitations of the study, and
conclusions drawn in the context of other studies.
This formula implies a linear flow in execution of the study, from aims to
methods, results, and conclusions—completely in keeping with the objec-
tivist approach to evaluation. Reporting an evaluation study using this
model encourages authors to be clear about the evaluation questions that
were addressed and the data that were used or collected to answer the
questions—helping the reader determine if the inferences drawn from the
data are justified.
Authors of papers describing studies of information resources should be
guided by the above structure, but may wish to add further sections or detail
within sections where necessary. For example, where novel statistical or
computational methods have been used, it is useful to include a paragraph
describing them in the methods section. In the case of measurement studies,
it is wise to include copies of the relevant instruments for publication as
figures or an appendix.
The above structure applies equally to reports of evaluation studies that
use the methods of randomized clinical trials. Because of the importance of
trials in providing the most credible objectivist evidence about the efficacy
of clinical interventions, 5 additional guidelines about reporting trials
have been published, including the work of the Consolidated Standards of
Reporting Trials group. 6 Some journals now require that all clinical trials
be reported according to these standards. This practice will aid groups such
as the Cochrane Collaboration, who are writing systematic reviews or meta-
analyses of the literature, by putting this literature into a more uniform and
directly comparable format. 7 Equally, because bibliographic systems
increasingly store the abstract along with the citation, many journals are
now requesting that authors structure the abstract of an article into sections