The primary author of this post is Dr. Tim Reilly.
A teacher might tell a white lie to students, to protect a student from nosey classmates. A cab driver might tell a white lie to reassure passengers in the car (“this is normal traffic”). In both of these cases, smoothing interactions over may be the best way to maintain trust and focus on the task at hand. For scientists, ‘smoothing things over’ may have much more serious consequences, such as misleading the public (e.g., on using alcohol), and harming others, for example with faulty medical products.
Concerns about scientific ethics are receiving more attention, especially amidst high profile examples of methodological favoritism (designing studies of alcohol use in consultation with alcohol producing companies) and even data fabrication, such as a study falsely claiming new uses for stem cell technology. Instances like these of dishonesty create obstacles for other researchers whose work hinges on the reliability of published research.
Discussions in science ethics tend to focus more on dishonesty than on a positive description of honesty. In hour-long conversations with scientists about what makes for good science and the qualities of a good scientist, data fabrication has been called, “the cardinal sin of science”. Still, this approach focuses scientists on negative morality, the lowest common denominator—not intentionally harming others. A parallel example can be found in consideration of the old moral mantra of Google, “Don’t be evil.”
What’s wrong with an emphasis on avoiding negative behavior? They ignore any responsibility to do good. Wouldn’t it be more satisfying to say being honest is a disposition of scientists, something that scientists seek to live out, rather than only avoiding dishonesty? Then honesty would be considered a central feature of scientific character rather than simply a descriptor of scientific behavior that is not intentionally dishonest.
Of course, scientific honesty includes avoiding dishonesty, even when dishonesty may be tempting, as when statistical significance is just over the borderline.
What does an honest disposition in science look like? It has positive components, such as pursuing accurate beliefs and understanding. These components rely on other pursuits. The honest scientist can be more confident that their beliefs are accurate if they are meticulous in their method, analysis, reporting, and interpretation of others’ work. In addition, the honest scientist reasons well and is appropriately skeptical of new findings, but open to adjusting their own views when its required, when examining data and methods of self and colleagues. Finally, a scientist can be honestly accountable through transparency, by sharing their data and methods clearly with others. Some examples may be helpful here.
Meticulousness. A scientist may misremember some of the experiments they conducted. They may have let cells culture for an extra day, for instance, and not have recorded it. This is a problem because more mature cells or cell colonies may function quite differently than younger ones, and so they find mistaken effects. If scientists don’t take meticulous notes, ensure the accuracy of their instruments, and follow the experimental procedure closely, they may not be aware that their results are driven by a mistake, and so could be unintentionally dishonest in reporting these findings. This would be a failure of meticulousness.
Reasoning and Skepticism. Even if the scientist is meticulous in this way, he or she may not consider important factors to account for in their research. As one example, consider evaluation of a program to help students form healthy relationships in the transition to college, especially with their roommates. A clear procedure is developed and followed for this study, with effective survey scales and a direct comparison between two schools with the same program, with half of the students at each school completing the program and half of the students not receiving the program. After completing this study, researchers may find that students at one university tend to be much more successful in building and maintaining friendships, regardless of whether they were involved in the program or not. What might be going on here? The researchers don’t have the evidence to say for certain. The researchers also didn’t consider or measure potentially relevant factors. Perhaps the university with students who fare better is a residential university, so students have more time, out of class, to build relationships. On the other hand, perhaps that university instead, for whatever reason, has more students from large families, who are used to living with roommates (siblings) and getting to know strangers (cousins from out of town). Regardless, even if these researchers are honest about what they found, they can’t tell the whole story, and may even be missing out on key parts. Good reasoning and skepticism would acknowledge these limitations, and perhaps anticipate them, gathering the relevant data to rule out alternative explanations. Such reasoning and skepticism are necessary to being honest in the way I am describing.
Transparent Reporting and Accountability. Taking the example further, transparent reporting and accountability may help those who have conducted a flawed study to recognize the flaws in their work, through the process of peer review. However, it is rarely, if ever, feasible to run a flawless study. Instead, scientists must, as one of those I interviewed described it “perform the least bad experiment” they can. The ‘perfect’ experiment may require technology, time, and resources that aren’t available.
Given these examples, considering honesty as something to be pursued helps to avoid these three possible pitfalls:
- intentional dishonesty (e.g., data fabrication or fraudulent analysis)
- unintentional dishonesty (e.g., see the misremembering example)
- insufficiently reasoned honesty can lead to flimsy and misleading findings (see the program comparison example)
Given this, it is heartening that my colleagues and I have found scientists’ conceptions of honesty include avoiding dishonesty, even when they find it tempting, and pursuing accurate beliefs through meticulousness, skepticism of new information, and sharing their own data and process transparently with others.
Thus, honesty is dependent on other dispositions which scientists must cultivate and maintain: meticulousness, reasoning, skepticism, and sharing transparently (which requires social risk). In my interviews, undergraduate researchers tend to mention that these are qualities they must actively seek to build, as they learn how to be a scientist. They are not natural qualities that one brings, fully formed, to science. Even if they were, they must be learned and applied in the particular field of science. Graduate students discuss the emphasis they place on making room for mistakes and other approaches to supporting undergraduate researchers in developing these dispositions, and so setting the stage for honesty.
In addition, in interviews some scientists offered cautionary tales about research groups fostering dishonesty, through a pressure for publication and a sense of academic competition. For instance, one experienced researcher described a laboratory in which the laboratory director put a researcher under such intense pressure to publish prematurely that the researcher felt the need to falsify an analysis to keep his job. This led to a fraudulent publication that was later discovered, ending the academic careers of both researchers.
One other feature of honesty mentioned by scientists has little to do directly with scientific experiments—honestly acknowledging the sources of one’s ideas and citing them accurately. Credit and recognition for scholarship are important currency in academic life, and not crediting others’ work (e.g., the controversy around Watson and Crick’s use of Rosalind Franklin’s data in the discovery of the double helix) or citing that work in misleading ways are forms of scientific dishonesty to be resisted.
These examples demonstrate that while honesty can be attributed to an individual (for instance the lab member in the final example could have refused to present falsified results and been fired), a robust consideration of scientists’ disposition toward honesty considers the context they are in and how honesty is cultivated and rewarded in it. In my interviews, it is clear scientists themselves don’t think of honesty as a trait that individuals possess or lack, but rather as a product of the formation of scientific character. Honesty is dynamic and as one achieves scientific status, learns about the social structure of science, and seeks to develop an accurate understanding of whatever phenomena being studied, an appreciation of the complex achievements of honest dispositions becomes apparent among senior scientists.
*Dr. Tim Reilly is assistant professor at Ave Maria University. This post was written under the auspices of the “Developing Virtue in the Practice of Science” grant project at the University of Notre Dame (directed by Celia Deane-Drummond, Darcia Narvaez, Thomas Stapleford) funded by the Templeton Religion Trust.