As always, we should start by preparing our brains for new learning. And what better way to do that than to ask some questions. Challenge your intuition with this quizz:
-
What percentage of scientists who participated in an anonymous survey of research practices at Dutch universities admitted to falsifying and/or fabricating data at least once between 2017 and 2020?
A. 8%
B. 10%
C. 70%
D. 39% -
According to the International Research Integrity Survey (IRIS), what is the most common questionable research practice (QRP) encountered by researchers based in Europe?
A. Data fabrication
B. Inadequate peer review
C. Poor supervision
D. Listing authors who did not contribute sufficiently -
What is the average placebo response in pain studies according to a meta-analysis of 114 placebo-controlled trials? A. 8%
B. 11%
C. 30%
D. 32% -
What is reanalysis in the context of scientific research?
A. The process of re-examining existing data or experiments with the goal of verifying or questioning the original findings.
B. The process of falsifying or fabricating data in scientific research.
C. The process of peer reviewing scientific research.
D. The process of listing authors who did not contribute enough in scientific research -
What is the most common type of research fraud in the United States according to a study published in the journal Accountability in Research?
A. Data fabrication
B. Inadequate peer review
C. Poor supervision
D. Gift authorship (adding researchers as co-authors without contributing much or any work to the paper)
You will find the correct answers below the article.
Fraud in science
Fraud in science, including data falsification and fabrication, is a significant problem in the research community. Studies have found that a significant proportion of researchers admit to engaging in fraudulent practices. For example,
an anonymous survey of research practices at Dutch universities found that 8% of scientists who participated in the survey confessed to falsifying and/or fabricating data at least once between 2017 and 2020.
More than 10% of medical and life-science researchers admitted to committing this type of fraud. The International Research Integrity Survey (IRIS) conducted a survey of around 47,000 academics in Europe and the United States about their experience of eight "questionable research practices" (QRPs).
The survey found that almost 70% of researchers based in Europe say that they have been involved in projects in the past three years that listed authors who did not contribute sufficiently to the work, making it the most common QRP encountered by the respondents. Inadequate peer review and poor supervision were the next most common QRPs.
Fraud may be less prevalent in psychotherapy research due to the lack of funding available for this type of research. This lack of funding may lead to fewer incentives for researchers to engage in fraudulent behavior. However, researcher allegiance in psychotherapy research has been found to account for a large portion of the variance in treatment outcomes, which may lead to more cautious interpretations of research findings. For example, Luborsky et al. (1999) reported that researcher allegiance in psychotherapy research has been found to account for 70% of the variance in the effect sizes of treatment comparisons. It seems reasonable to agree with Messer and Wampold (2002), who state “How odd it is, then, that we continue to examine the effect of different treatments (accounting for less than 1% of the variance) when a factor such as the allegiance of the researcher accounts for nearly 70% of the variance!” (p. 23).
A new study published in the journal Accountability in Research has found that 'gift' authorship, where researchers are added as co-authors without contributing much or any work to the paper, is the most common type of research fraud in the United States. The study surveyed 613 researchers from 100 leading research-intensive universities in the US and found that data fabrication was the least common type of research fraud. Gift authorship can take many forms, such as adding famous scholars to the author list to increase the chances of being published, or 'honorary authorship' where senior researchers are named as authors because of their stature within the institution or because they helped to obtain the funding. The study suggests that graduate students, particularly those from low- and middle-income countries, often do not receive adequate training on the ethics of publishing and authorship practices.
When one adds findings like these to the equation, one may become worried about the value of science:
-A study by Nature in 2016 found that of the 1,576 papers it had published since 2010, only 39% were able to be replicated successfully.
-Another study by Science found that of the 100 papers it had published since 2015, only 36% were able to be replicated successfully.
-A study by the National Institutes of Health (NIH) in 2012 found that of the 30,000 papers published in the field of preclinical research, only 11% had been successfully replicated.
-A meta-analysis of 114 placebo-controlled trials found that the average placebo response in pain studies was 30%, while a systematic review of antidepressant trials found that the average placebo response was 32%.
In addition to the issues of reproducibility and replicability, reanalysis is another important aspect of ensuring the integrity of scientific research. Reanalysis refers to the process of re-examining existing data or experiments with the goal of verifying or questioning the original findings. This process can be crucial in identifying errors or inconsistencies in the original research and can help to improve the overall quality and reliability of the scientific literature.
One of the challenges of reanalysis is that it can be time-consuming and resource-intensive, and it often requires specialized expertise. This can make it difficult for researchers to conduct reanalysis on their own, especially if they are not familiar with the original study or the methods used.
Despite these challenges, there are a growing number of initiatives and resources aimed at promoting reanalysis and improving the reproducibility of scientific research. For example, the Data Reuse and Reproducibility Network (DR2) is a collaboration of researchers, journals, funders, and other organizations that aims to promote data sharing and reanalysis in the scientific community.
In addition, many journals now have guidelines and policies in place to encourage and support reanalysis, and some even have dedicated reanalysis sections or special issues to showcase reanalysis studies.
However, it is important to note that reproducibility and replicability are not the only ways to ensure the validity and integrity of scientific research. Reanalysis, the process of independently reviewing and analyzing previously collected data, is a crucial step in verifying the reliability of research findings. Reanalysis allows for the detection of errors, inconsistencies, and potential fraud in previously published studies, and can also lead to the discovery of new insights and discoveries. However, reanalysis is often overlooked and undervalued in the scientific community, despite its potential to improve the overall quality and reliability of scientific research. To combat issues of fraud and irreproducibility, it is important for researchers to prioritize not only reproducibility and replicability, but also reanalysis in their research practices.
It is important to note that while fraud and misconduct are serious problems, they are relatively rare. The vast majority of research is conducted ethically and with integrity. Additionally, many organizations and journals have implemented strict guidelines and procedures to detect and prevent fraud and misconduct. However, there are also issues with reproducibility and replicability of results, which can be attributed to factors such as poor experimental design, lack of transparency in data and methods, and publication bias.
Correct answers
A. 8%
B. Listing authors who did not contribute sufficiently
C. 30%
A. The process of re-examining existing data or experiments with the goal of verifying or questioning the original findings.
D. Gift authorship (adding researchers as co-authors without contributing much or any work to the paper)
Sources
https://www.nature.com/articles/d41586-023-00016-1
https://www.nature.com/articles/d41586-021-02035-2
https://www.nature.com/articles/d41586-020-02847-8
https://www.nature.com/articles/525025a
https://www.nature.com/nature-index/news-blog/gift-ghost-authorship-what-researchers-need-to-know
Messer, S. B., & Wampold, B. E. (2002). Let's face facts: Common factors are more potent than specific therapy ingredients. Clinical Psychology-Science and Practice, 9(1), 21-25. https://doi.org/DOI 10.1093/clipsy/9.1.21
