This account of bad science is meant to cover the whole range of scientific
methods, from statistical analysis of large numerical data sets to qualitative
research interviews. Translated into practical terms, some obvious recommenda-
tions come out:
. The study should start with a satisfactory literature review which permits the
definition of the research question, in such way as to show that the question is
important, it has practical relevance and we don't already know the answer to
the question. No inquiry is so naõÈve 'ornovel 'that it does not build in some
way on previous work or on previously developed methods, and these debts
need to be brought into view and analysed, so far as this is possible.
. The design of the study must be reliable and likely to answer the research
question in such a way that the validity of the answer is determinate and the
findings of the study are interpretable and applicable by other practitioners and
researchers.
. The results of the study must be publishable and actually published, within a
reasonable time from the completion of the study, even if negative, to permit
other researchers and the public to learn from the study 'including its weak-
nesses no less than its strengths). The publication should be a fair and accurate
account of the research design and results. There is an equivalent duty on the
editor of the journal or book and reviewers for the journal or book to give a fair
and competent assessment of the article or chapter submitted for publication.
All of these recommendations are now included in the Declaration of Helsinki,
which is the most important international ethical guideline regulating biomedical
research. However, they are here restated in language which shows their applic-
ability as widely as possible to the diversity of research methods used in nursing
today, including qualitative and health services research methods.
Having defined bad science', it should be clear whybad science is bad ethics'.
In the first place, research involves exposing patients or colleagues or other
research subjects to the risks of the research. Hence, if this research is unlikely to
produce reliable results, it is arguable that the subjects are exposed to risk without
this in any way being balanced by the prospect of benefit to society. To the extent
that participants are taking part with altruistic motives, bad research mis-
represents itself as an opportunity to benefit others, when it has no prospect of
doing so. As such, it could be seen both as an insult to the altruism of the parti-
cipants, and as deception of them.
To the extent that the research offers some benefit to the participants in terms of
access to new treatment, increased access to nursing or other health care services,
or financial or other inducements, there is still an issue about the waste of
resources bad research involves. Research always involves staff time and use of
basic resources, even where there is no additional grant funding component.
Hence, there is always an `opportunity cost', as the economists say, involved in
doing research. The opportunity cost of bad research is at least the opportunity of
using staff and other resources more effectively, either in caring for patients or
carrying out bona fide research, for instance. Research ethics typically ignores the
ethical issues involved in resources and facilities management in the health
services, but this is morally shortsighted.
ClinicalResearchandPatients 281