Are you in Canada? Click here to proceed to the HK Canada website.

For all other locations, click here to continue to the HK US website.

Human Kinetics Logo

Purchase Courses or Access Digital Products

If you are looking to purchase online videos, online courses or to access previously purchased digital products please press continue.

Mare Nostrum Logo

Purchase Print Products or eBooks

Human Kinetics print books and eBooks are now distributed by Mare Nostrum, throughout the UK, Europe, Africa and Middle East, delivered to you from their warehouse. Please visit our new UK website to purchase Human Kinetics printed or eBooks.

Feedback Icon Feedback Get $15 Off

Seven areas of research misconduct

This is an excerpt from Research Methods in Physical Activity-8th Edition by Jerry R. Thomas,Philip E. Martin,Jennifer L. Etnier & Stephen J. Silverman.

The U.S. Department of Health and Human Services (HHS) Office of Research Integrity (ORI) has defined research misconduct for all U.S. agencies:

Research misconduct is fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results. (Federal Register, 42 CFR Part 93, May 17, 2005) It does not include honest error or differences of opinion.

Shore (1991) identified seven areas in which research misconduct might occur; each is discussed in the following subsections. A 1993 issue of Quest (Thomas & Gill) included several thought-provoking articles on ethics in the study of physical activity. Steneck (2007) also is a valuable resource on the topic. For more resources on the responsible conduct of research (RCR), visit the website of the HHS ORI (


Plagiarism is the act of using someone else’s ideas, writings, or drawings. Of course, this is completely unacceptable in the research process (including writing). Citing a reference or source provides credit for ideas and information, but it does not give credit for words or phrasing. Even when a citation to information from a source is provided, it is not appropriate to use small, medium, or large sections of text from that source. Even changing a few words here or there while maintaining a source’s basic sentence structure is insufficient. Directly quoting the words of someone else combined with a reference to the source is one approach to give appropriate credit, but it should be used sparingly. Thus, developing the skill to paraphrase content from other sources is important to avoid plagiarism. Plagiarism carries severe penalties at all institutions. A researcher who is publicly outed for plagiarism carries a lifelong stigma in their profession. No potential outcome is worth the risk involved.

On occasion, a graduate student or faculty member can inadvertently be involved in plagiarism. This situation generally occurs on work that is coauthored. If one author plagiarizes material, the other could be equally punished despite being unaware of the plagiarism. Although no means of protection is surefire (except not working with anyone else), never allow a paper with your name on it to be submitted (or revised) unless you have reviewed and approved the complete paper in its final form.

In scientific writing, originality is also important. Common practice is to circulate manuscripts and drafts of papers among scholars, including graduate students, who are known to be working in a specific area. If ideas, methods, findings, and so on are borrowed from them, proper credit should always be given.

Data Fabrication and Falsification

Based on information in “Publications Output: U.S. Trends and International Comparisons” reported by the National Science Foundation in 2019 (, articles published in peer-reviewed science and engineering journals increased from 1.8 million (2008) to 2.6 million (2018). Given this huge volume of research, it is not surprising to learn that scientists have occasionally been caught making up or altering research data. Of course, this action is completely unethical, and severe penalties are imposed on people who are caught. Pressure has been particularly intense in medical and health-related research because such research is often expensive, requires external funding, and involves risk. The temptation is great to make a little change here or there or to make up data because “I only need a few more participants, but I am running out of time.” The odds of being detected in these types of actions are high, but even if you should get away with it, you will always know that you did it, and you will probably put other people at risk because of your actions.

Although graduate students and faculty may knowingly produce fraudulent research, established scholars are sometimes indirectly involved in scientific misconduct. This may occur when they work with other scientists who produce fraudulent data that follow the predicted outcomes (e.g., as in a grant-funded proposal that suggests which outcomes are probable). In these instances, the established scholar sees exactly what they expect to see in the data. Because this result verifies the hypotheses, the data are assumed to be acceptable. One of the most famous cases involved Nobel laureate David Baltimore, who coauthored a paper with principal authors Thereza Imanishi-Kari and David Weaver that was published in Cell in April 1986 (Weaver et al.). Baltimore checked the findings, but he saw in the data the expected outcomes and agreed to submit the paper. The fact that the data were not accurate subsequently led to Baltimore’s resignation as president of Rockefeller University. Thus, although Baltimore was not the principal author of the paper, his career was seriously damaged.

Falsification can also occur with cited literature. Graduate students should be careful how they interpret what an author says. Work of other authors should not be “bent” to fit projected hypotheses. For that reason, graduate students should read original sources instead of relying on the interpretations of others, because those interpretations may not follow the original source closely.

Nonpublication of Data

Nonpublication of data refers to the exclusion of a selected subset of data from a study because they do not support the desired outcome. This tactic is sometimes known as “cooking” data. A thin line separates the action of eliminating “bad” data from “cooking” data. Bad data should be caught, if possible, at the time of data acquisition. For example, if a test value seems too large or small and the researcher checks the instrument and finds it out of calibration, eliminating this bad data is good research practice. But deciding that a value is inappropriate when data are being analyzed and then changing the value is cooking the data.

Another term applied to unusual data is outlier. Some people have called such data outright liars, suggesting that the data are bad. But extreme values now are sometimes trimmed. Just because a score is extreme does not mean that it is based on bad data. Although extreme scores can create problems in data analysis, trimming them automatically is a poor practice (see chapter 10 on nonparametric analysis for one solution to this problem). A researcher should always carefully explore potential causes of a participant’s extreme response or outcome and have strong justification for declaring an outlier and eliminating data from further analysis.

Perhaps the most drastic example of nonpublication of data is the failure to publish results that do not support projected hypotheses. Journals are often accused of a publication bias, meaning that only significant results are published. But authors should publish the outcomes of solid research no matter what findings support projected hypotheses. The results from well-planned studies based on theory and previous empirical data have important meaning regardless of whether the predicted outcomes are found.

Faulty Data-Gathering Procedures

A number of unethical activities can occur at the data collection stage of a research project. In particular, graduate students should be aware of issues such as these:

  • Continuing with data collection from participants who are not meeting the requirement of the research (e.g., poor effort; failure to adhere to agreements about diet, exercise, rest)
  • Using malfunctioning equipment
  • Treating participants inappropriately (e.g., failing to follow the procedures and guidelines agreed to by the investigator and the human subjects committee)
  • Recording data incorrectly

For example, a doctoral student whom we know was collecting data on running economy in a field setting. Participants returned several times to be video-recorded while repeating a run at varying stride lengths and rates. On the third day of testing, a male runner performed erratically during the run. When the experimenter questioned him, she learned that he was hung over from drinking with his buddies the previous evening. Of course, she wisely discarded the data and scheduled him for another run several days later. Had she not noted the unusual nature of his performance and questioned him carefully, she would have included data with skewed results because the participant was not adhering to previously agreed-upon conditions of the study.

Poor Data Storage and Retention

Based on a survey of authors for 516 biology-related journal articles, Vines and colleagues (2014) estimated that data availability declined 17% per year. The primary reasons for data unavailability were the data had either been lost or stored on media or in formats that were no longer accessible. Data must be stored and maintained as originally recorded and not altered. Lab notebooks documenting data collection and analysis should be stored in a secure and safe location. All original records should be maintained so that the original data are available for examination. Federal agencies and most journals suggest that original data be maintained for at least three years following publication of the results. We recommend that data be maintained indefinitely.

Data ownership is a related and equally important issue. A graduate student who expends countless hours collecting and analyzing data for a thesis or dissertation project might assume that they own those data. Ownership, however, typically belongs to the institution where the data were collected or, in the case of funded research through grants or contracts, to the funding agency. Government funding agencies usually allow the research institution to manage data, whereas private companies are more likely to retain the right to all data generated from contract funding (Steneck, 2007). A graduate student who wishes to take research data with them when leaving an institution should plan to leave the original data with their advisor or department and seek permission of their advisor to retain a copy of the data.

Misleading Authorship

A major ethical issue among researchers involves joint research projects or, more specifically, the publication and presentation of joint research efforts. Generally, the order of authorship for presentations and publications should be based on the researchers’ contributions to the project. The first author is usually the researcher who developed the idea and the plan for the research. Other authors are normally listed in the order of their contributions (see Fine & Kurdek, 1993, for a detailed discussion and case studies). Some fields also recognize contributions by a “senior author,” typically a well-established investigator closely involved in the research who appears last among the authors of a research article. It is not uncommon for the advisor or supervisor of a graduate student’s research to serve as a senior author.

Although establishing whether a person warrants authorship credit and determining author order sounds easy enough, authorship decisions are difficult and can be contentious at times. Sometimes researchers make equal contributions and decide to flip a coin to determine who is listed first. Regardless, decide the order of authors at the beginning of a collaborative effort. This approach saves hard feelings later, when everyone may not agree on whose contribution was most important. If the contributions of various authors change over the course of the research project, discuss a change at that time.

A second issue is who should be an author (see Crase & Rosato, 1992; Grossman & DeVries, 2019; and Venkatraman, 2010, for good discussions of authorship). Studies occasionally have more authors than participants. In fact, sometimes the authors are also the participants. When you look at what participants must go through in some research studies, you can see why only a major professor’s graduate students would allow such things to be done to them. Even then, they insist on being listed as authors as a reward. More seriously, the following two rules should help define authorship:

  • Authorship should involve only those who contribute directly to the research project. According to Steneck (2007, p. 134),

authorship is generally limited to individuals who make significant contributions to the work that is reported. This includes anyone who: was intimately involved in the conception and design of the research, assumed responsibility for data collection and interpretation, participated in drafting the publication, and approved the final version of the publication.
Many journals and professional organizations have established similar, but not identical, criteria for authorship credit. This listing does not necessarily include the laboratory director or a graduate student’s major professor. The only thing that we advocate in the chain letter in the Chain Letter to Increase Publications sidebar is the humor.

  • Technicians do not necessarily become joint authors. Graduate students sometimes think that because they collect the data, they should be coauthors. Only when graduate students contribute to the planning, analysis, and writing of the research report are they entitled to be listed as coauthors. Even this rule does not apply to grants that pay graduate students for their work. Good major professors involve their graduate students in all aspects of their research programs; thus, these students frequently serve as coauthors and technicians.

Unacceptable Publication Practices

Another scientific dishonesty concern deals with coauthored or joint publications—specifically, those between the major professor and the graduate student. Major professors do (and should) immediately begin to involve graduate students in their research programs. When this happens, the general guidelines that we suggested earlier apply. But two conflicting forces are at work. A professor’s job is to foster and develop students’ scholarly ability. At the same time, pressure is increasing on faculty to publish so that they can obtain the benefits of promotions, tenure, outside funding, and merit pay. Being the first (senior) author is a benefit in these endeavors. As a result, faculty members want to be selfless and assist students, but they also feel the pressure to publish. This issue may not affect senior faculty, but it is certainly significant for untenured assistant professors. As mentioned previously, there are no hard-and-fast rules other than that everyone agree before the research is undertaken.

The thesis or dissertation is a special case. By definition, the thesis or dissertation is how a graduate student demonstrates competence to receive a degree. Frequently, for the master’s thesis, the major professor supplies the idea, design, and much of the writing and editing. In spite of this, we believe that it should be regarded as the student’s work. In other words, the student should be first author on any publication emanating from the thesis research. The dissertation should always be regarded as the student’s work, but second authorship for the major professor on either the thesis or the dissertation is acceptable under certain circumstances. The American Psychological Association's Ethical Principles of Psychologists and Code of Conduct (2017, section 8.12) has defined these circumstances adequately, and we recommend the use of their guidelines:

  • Only second authorship is acceptable for the dissertation supervisor.
  • Second authorship may be considered obligatory if the supervisor designates the primary variables, makes major interpretive contributions, or provides the database.
  • Second authorship is a courtesy if the supervisor designates the general area of concern, is substantially involved in the development of the design and measurement procedures, or substantially contributes to the writing of the published report.
  • Second authorship is not acceptable if the supervisor provides only encouragement, physical facilities, financial support, critiques, or editorial contributions.
  • In all instances, agreement should be reviewed before the writing for publication is undertaken and at the time of the submission. If disagreements arise, they should be resolved by a third party using these guidelines.

Authors must also be careful about dual publication. Sometimes, this circumstance is legitimate; for example, a scientific paper published by one journal may be reprinted by another journal or in a book of readings (this should always be noted). Authors may not, however, publish the same paper in more than one copyrighted original research journal. But what constitutes the “same paper”? Can more than one paper be written from the same set of data? The line is rather hazy. For example, Thomas (1986, pp. iv-v) indicated that

frequently, new insights may be gained by evaluating previously reported data from a different perspective. However, reports of this type are always classed as research notes whether the reanalysis is undertaken by the original author or someone else. This does not mean that reports which use data from a number of studies (e.g., meta-analyses, power analyses) are classed as research notes.

Generally, good scientific practice is to publish all the appropriate data in a single primary publication. For example, if both psychological and physiological data were collected as a result of a specific experiment on training, publishing these separately may not be appropriate. Frequently, the main finding of interest is the interaction between psychological and physiological responses. But in other cases, the volume of data may be so large as to prohibit an inclusive paper. Sometimes the papers can be published as a series; at other times they may be completely separate. Another example is a large-scale study in which a tremendous amount of information is collected (e.g., exercise epidemiological or pedagogical studies). Usually, data are selected from computer records (or videos) to answer a set of related questions for a research report. Researchers may then use a different part (or even an overlapping part) of the data set to address another set of questions. More than one legitimate publication may then result from the same data set, but the authors should identify that more than one paper has been produced. Researchers who do not follow these general types of rules may be viewed as lacking scientific objectivity in their work and certainly as lacking in modesty (Gastel & Day, 2016).

Most research journals require that the author include a statement that the paper has not been previously published or submitted elsewhere while the journal is considering it. Papers published in one language may not be published as an original paper in a second language.

Chain Letter to Increase Publications

Dear Colleague:

We are sure that you are aware of the importance of publications in establishing yourself and procuring grants, awards, and well-paying academic positions or chairpersonships. We have devised a way your curriculum vitae can be greatly enhanced with very little effort.

This letter contains a list of names and addresses. Include the top two names as coauthors on your next scholarly paper. Then remove the top name and place your own name at the bottom of the list. Send the revised letter to five colleagues.

If these instructions are followed, by the time your name reaches the top of the list, you will have claim to coauthorship of 15,625 refereed publications. If you break this chain, your next 10 papers will be rejected as lacking in relevance to real-world behavior. Thus, you will be labeled as ecologically invalid by your peers.


Jerry R. Thomas, Professor

Jack K. Nelson, Professor

List as coauthors:

Jerry R. Thomas

Jack K. Nelson

I.M. Published

U.R. Tenured

C.D. Raise

More Excerpts From Research Methods in Physical Activity 8th Edition



Get the latest insights with regular newsletters, plus periodic product information and special insider offers.