Data publication and reporting is the process of preparing and disseminating research findings to the scientific community. Scholarly disciplines can only advance through dissemination and review of research findings at professional meetings and publications in discipline-related journals. The tacit assumption in publishing is one of trust between the author(s) and readers regarding the accuracy and truthfulness of any submission.

The practice of ensuring research integrity is relevant at all stages of research investigation, from early conceptualization, design, implementation, to analysis. This practice also extends to the stage of documenting and preparing results for publication. In this process, researchers may experience many more challenges to preserving research integrity.

Considerations/issues in data reporting and publishing

There are often factors in research settings that can result in compromises to data integrity. These factors may facilitate conditions where the goal of conducting research in as objective a manner as possible can sometimes be challenged. These can be categorized as either external or internal factors as follows:

External Factors:

Internal Factors:

Importance of accurate and honest data reporting

Investigators demonstrating lapses of integrity while engaged in data reporting and publishing can have a negative influence in the direction of future research efforts, threaten to compromise the credibility of a particular field of study, and may ultimately risk the well-being and safety of the public in general, as well as research subjects in particular.

Sources of guidance promoting good data reporting practices and publishing include faculty advisors who carefully instruct graduate students, departmental chairpersons mentoring researchers new to the field, regular review of published university policies, existing codes of professional ethics, or established government rules and regulations. Deficiencies in training or a lack of awareness of existing policies, codes, or rules may increase the likelihood of a deviation from the acceptable standards of practice in reporting and publishing.

Listed below are some issues related to integrity of data reporting and publication:

Misrepresentation

Due to problems data in collection, researchers may omit data that is not supportive of the research hypothesis. Alternately, data may be fabricated if the data collection process was somehow interrupted or data was lost, and the researchers believe the invented data would have been similar to what was anticipated. In either case, the true scope of the data findings remains hidden from readers who are unable to accurately assess the validity of the findings.

Plagiarism

Plagiarism is the act of taking credit for ideas or data that rightfully belongs to others. Related to this is the theft of ideas from grants and drafts of papers that a researcher has reviewed. This harms the researcher(s) from which the idea(s) or data was appropriated improperly acknowledged.

Selectivity of reporting / failure to report all pertinent data

This is the practice of only using data that supports one’s research hypothesis and ignoring or omitting data that does not. A related practice is inaccurate reporting of missing data points. As explained under “Misrepresentation” earlier, the true scope of the data findings remains hidden from readers who are unable to accurately assess the validity of the findings.

Failure to disclose conflicts of interest

Editors, reviewers, or readers who are not aware of possible conflicts of interest (financial and otherwise) may not have an opportunity to adequately assess the validity of research findings without being aware of possible undue influences from the sponsors of an investigation. These conflicts may compromise researchers’ credibility in their fields.

Publication bias / neglecting negative results

Since the vast majority of research findings submitted to professional journals tend to be ‘positive’ in nature, the literature in most scientific fields demonstrates a negative bias. This in part reflects the reluctance of journal editors to publish articles with negative findings. Thus, researchers are less willing to report findings that fail to demonstrate an intended effect or yield an expected result. The value of these publications could be substantial in that other investigators would not needlessly pursue a fruitless path of research.

Analysis of data by several methods to find a significant result

This is also known as ‘milking’ or ‘dredging the data’ and involves researchers utilizing a variety of statistical tests in the hopes of yielding a significant result. The proper procedure would be to base the selection of desired tests on a theory or theoretical framework rather than selecting tests a priori. Other related statistical issues include reporting percentages rather than absolute numbers due to small sample size, reporting differences when statistical significance is not reached suggesting a certain trend exists, reporting no difference when statistical power is inadequate, and failure to include the total number of eligible participants. The importance of this last point is the difficulty for readers to be able to determine whether a dismal non-respondent rate might compromise the representativeness of respondents.

Inadequate evaluation of prior research

This refers to an insufficient review of available literature that presents an incomplete picture of the current status of a particular research area. A critique of the included citations may lack the required depth of analysis and fail to justify the need for proposed research.

Ignoring citations or prior work that challenge stated conclusions or call current findings into question

Selective inclusion of citations that minimize threats to the justification for the present study can compromise the integrity of the study. Whether done intentionally or not, omissions can have the untoward consequence of providing support for an author’s position.

Misleading discussion of observations

This may result from using inappropriate statistical tests, neglecting negative results, omitting missing data points, failing to report actual numbers of eligible subjects, using inappropriate graph labels or terminology, and data dredging. These can result in readers becoming less able to objectively critique the findings.

Reporting conclusions that are not supported

Faulty data collection, inappropriate analyses, gaps in logic, and unexplained deviation from conventionally accepted methods of interpretation can result in conclusions that are not valid. Readers cannot assess the validity of the conclusions for themselves unless all the necessary information is honestly reported.

Breaking down of a single piece of research into multiple overlapping reports

This can occur when the distinction and differences in findings between reports is negligible and the focus is publishing for quantity versus quality. A related practice is submission of duplicate publications in journal from different disciplines or in different languages. The expectation is that investigators would not read journal from different fields of study or languages. Literature reviews or meta-analyses that are conducted may lead to an inaccurate assessment of findings from a particular research area due to duplicate publications of the same study in different journals.

Just Attribution of Authorship

Publication disputes generally fall into four categories(Ritter, Washington, 2001):

1. a researcher is listed as an author but did not have a chance to review or approve the manuscript
2. a researcher is promised first authorship when the project is completed, but the principal investigator adds the work of someone else, who then becomes first author
3. a researcher claims first authorship on the basis of the amount of work he or she did when not given that recognition, and
4. after leaving a laboratory, a researcher does not receive credit in an article that includes his or her work. Related to this is submission of manuscripts not seen and reviewed by all the listed co-authors of a publication

A fair and equitable understanding of each author’s contribution to published research provides clear credit and acknowledgement for advancing a field of study.

Inappropriate use of terminology without precise definitions

A potential barrier to successful cross-disciplinary investigations is the use of field-specific terminology. Encouraging the use of precise definitions can reduce confusion and promote understanding of research conducted.

Inflation of research results for the media

This involves providing statements for public and not professional consumption that are insufficiently supported by data for the purpose of publishing un-reviewed or untested results in a non-scientific or non-scholarly magazine/media. Premature reporting of results that turn out to be unsubstantiated may compromise the credibility of a particular field.

Publishing in peer-reviewed journals or presenting in scholarly meetings is the primary mechanism for investigators to disseminate their findings to the research community. This community relies on authors(s) to report the events of a study honestly and accurately. All researchers should be aware of the issues that compromise the integrity of data reporting and publishing. Ensuring integrity is essential to promoting the credibility of all fields of study.

Ethics Resources:

http://pubs.acs.org/cen/topstory/7946/7946sci1.html

http://onlineethics.org/reseth/mod/auth.html

References:

Marco, C.A., Larkin, G.L. (2000) Research ethics: ethical issues of data reporting and the quest for authenticity. Acad Emerg Med (Academic emergency medicine: official journal of the Society for Academic Emergency Medicine.), 7 (6): 691-694.

Price, J.H., Drake, J.A., Islam, R. (2001). Selected ethical issues in research and publication: perceptions of health education faculty. Health Education & Behavior, 28 (1): 51-64.

Stephen, K.R., Washington, C., Washington, E.N. (2001). Publication ethics: rights and wrongs: Balancing obligations and interests surrounding dissemination of research is an arduous task. Science & Technology , 79 (46): 24-31.

	Top