Deficient epistemic virtues and prevalence of epistemic vices as precursors to transgressions in research misconduct

Abstract

Scientific research is supposed to acquire or generate knowledge, but such a purpose would be severely undermined by instances of research misconduct (RM) and questionable research practices (QRP). RM and QRP are often framed in terms of moral transgressions by individuals (bad apples) whose aberrant acts could be made conducive by shortcomings in regulatory measures of organizations or institutions (bad barrels). This notion presupposes, to an extent, that the erring parties know exactly what they are doing is wrong and morally culpable, but had nonetheless proceeded to commit wrongful acts. However, a confession of intent to deceived is often not readily admitted by perpetrators of RM. I posit that beyond the simplistic notion of conscious moral transgression, deficits in epistemic virtues and/or the prevalence of epistemic vices have important roles to play in initiating and driving RM/QRP. For the individual perpetrator, deficits in epistemic virtues could lead to or amplify errors in one’s desperate attempt to be accomplished or to excel, and pushes one across the ethical line or down the slippery slope of misconduct. Likewise, a lack of epistemic virtue within perpetrators’ institution or organization could make it conducive for deceitful acts and suppress indications and warning signs for the former. Furthermore, epistemic vices exhibited by reviewers, editors and journals could also promote RM/QRP. In this view, epistemic failings, rather than widespread moral deficiencies of individuals within the research ecosystem, may underlie the prevalence of RM/QRP.

Introduction

Honesty, a pivotal moral virtue, is one that is most valued in conducting and reporting scientific research (for that matter, in all fields and settings of research). Research misconduct (RM) occurs when researchers act deceitfully or in a manner that violate acceptable norms, such as fabrication, falsification, or plagiarism (FFP) of research proposals, data or results (Gross, 2016; US Office of Research Integrity, 2022). ¹ There are also a number of research malpractices that are non-FFP in nature, such as p-hacking, hypothesizing after results are known (HARKing) and selective reporting, that are collectively termed questionable research practices (QRPs) (Andrade 2021; Fiedler and Schwarz, 2015). Both RM and QRPs are disturbingly common (de Vrieze 2021; Xie et al., 2021), and both contribute to irreproducibility in research (Baker, 2016; Saltelli and Funtowicz, 2017), the latter a phenomenon of crisis proportions that could potentially undermine public trust and societal support for research.

A conventional way of looking at the prevalence of RM and QRPs is through the lenses of morality and expectations of integrity on the part of the researcher. An individual who errs morally in research, be it one yearning for quick success, stressed by pressures to perform, or otherwise motivated to cheat, would be a prototypical “bad apple” (Kish-Gephart et al., 2010; Redman, 2013). Findings have also indicated that the sociological settings and systemic or institutional factors might made it conducive for acts of misconduct to occur, thus providing the context of a “bad barrel” view (Committee on Science Engineering Medicine and Public Policy, and Committee on Responsible Science, 2017; Davis et al., 2007; Kish-Gephart et al., 2010). Looking at RM from a moral perspective presupposes, or emphasizes, the occurrence of moral transgressions on the part of perpetrators regardless of the motivating or facilitating factors. In other words, a perpetrator is perceived to know clearly (or is supposed to know) what they are doing is wrong and morally culpable, but has nonetheless proceeded to commit RM.

However, in misconduct investigations, a confession of intent to deceive is often not readily forthcoming, even when evidence of misconduct is clear. Many acts of misconduct are also clumsy or somewhat awkward, with apparently little concern on the part of the perpetrator to shrewdly avoid attention or detection. Also, perpetrators often cite excuses for their acts that would attest to their ignorance, folly or circumstantial lapse of judgement that belie their academic credentials and years of training, rather than any intent to cheat. While these excuses may simply reflect attempts in grasping at straws, some perpetrators do appear to have deeper-lying deficits beyond an apparent moral breakdown. They might indeed not know, or have forgotten, how knowledge should be acquired.

The main purpose of scientific research is to acquire or generate knowledge in a focused and intensive manner. Scientific research is carried out by humans, and the inherently human nature of scientific inquiry makes its activities, products, as well as misadventures dependent on epistemic values. A concept in epistemology that is important for research is that of epistemic virtue or its opposite, epistemic vice (Paternotte and Ivanova, 2017). Epistemic virtues include not only truth-conducive cognitive abilities such as reliable perception and good reasoning, but importantly also traits such as intellectual honesty/integrity, courage, humility, open-mindedness and impartiality (Baehr, 2011). Epistemic vices include negative intellectual traits such as gullibility, dogmatism, prejudice, closed-mindedness, negligence, indifference to truth, and rigidity (Cassam, 2016, 2019; Meyer et al., 2021a, b). The importance of intellectual or epistemic virtues (Turri et al., 2018) in scientific research appears intuitively clear, and has been the subject of several recent analyses (Mendoza-De Los Santos, 2023; Paternotte and Ivanova, 2017; Uygun-Tunç and Pritchard, 2022).

In this article, I suggest that deficiencies in epistemic virtues, and/or a prevalence of epistemic vices, would contribute significantly to RM/QRP, with the latter being driven at times, if not primarily, by epistemic rather than moral failings. ² This epistemic virtue deficiencies could occur on the part of not just the individual RM perpetrator, but also the hosting institution and those in academic publishing (reviewers, editors, and journals). In the sections that follow, I shall first looked briefly at epistemic virtues/vices as defined by Cassam and others and their relevance to scientific research (Baird and Calvard, 2019; Cassam, 2016; Meyer et al., 2021a, b; Uygun-Tunç and Pritchard 2022). I shall then examine how deficiencies in epistemic virtue and the prevalence of epistemic vices could lead an individual, an institution or components within a publishing entity down the slippery slope to RM with the aid of fictitious case vignettes. Further, I also propose some ways of assessing and mitigating epistemic vices’ influence on RM/QRP.

Epistemic virtues, vices and their importance in scientific research

Virtues are generally defined by good moral character and epistemic virtues (or intellectual virtues) would indicate analogously positive traits in acquiring knowledge. An epistemically virtuous person is someone who is determined to seek truth and knowledge without being swayed by self-interest or feelings (Baril, 2016). Epistemic vices, on the other hand, would be character traits, attitudes or thinking styles that systematically get in the way of knowledge acquisition (Cassam, 2016, 2019). Some of these traits would include gullibility, dogmatism, prejudice, closed-mindedness, negligence as mentioned by Cassam (2016), as well as “indifference to truth” and “rigidity”, as expounded upon by Meyer et al. (Meyer et al., 2021a, b). These traits can be observed generally, and not peculiar to scientific activities.

Several authors have also recently elaborated on the relevance of epistemic virtues and vices specifically in scientific endeavors. Paternotte and Ivanova noted that epistemic virtues have a positive effect on science because they accelerate successful convergence amongst scientists in theory choice situations (Paternotte and Ivanova, 2017). Mendoza-De Los Santos has analyzed how intellectual virtues are linked to scientific research tasks and those of vigilance and intervention in the uses of scientific knowledge, and concluded that intellectual virtues are key to achieving higher quality scientific research and adequate foresight on the uses and impacts of scientific findings (Mendoza-De Los Santos, 2023). The field of psychological and behavioral sciences has a credibility crisis as many celebrated or key findings cannot not subsequently be replicated (Malich and Munafò, 2022; Open Science Collaboration, 2015). Indeed, reproducibility has been an issue for many fields in the natural and social sciences (Baker, 2016). In relation to science’s credibility crisis, Uygun-Tunç and Pritchard noted that while individual virtue is neither necessary nor sufficient for science to be successful, the presence of a “significant proportion of epistemically virtuous scientists in a scientific community” is a necessary condition for collective epistemic success in a scientific community (Uygun-Tunc and Pritchard, 2022). On the other hand, Hoekstra and Vazire focused on the often ignored importance of intellectual humility in science and credibility-enhancing reforms, and provided a set of recommendations on how to be intellectual humble in presenting research articles (Hoekstra and Vazire, 2021). These analyses and thoughts collectively attest to the importance of epistemic virtue in science.

Epistemic virtue in science would comprise of aspects of professional competency and responsibility. Uygun-Tunc and Pritchard (2022) noted that an epistemically virtuous scientist can be construed as “. . . someone who competently, and responsibly, contributes to the advancement of scientific knowledge,” and who “. . . puts epistemic goals and values above all else in their approach to inquiry” (Uygun-Tunc and Pritchard, 2022). However, these scientists need not have exclusively epistemic goals in their research endeavors, and could also be motivated by non-epistemic goals such as recognition and career advancement. Interestingly, the authors do not expect an epistemically virtuous scientist to also be exceptionally morally virtuous, and suggested that “. . . epistemic virtue is perfectly compatible with moral vice or other failings of character.” On the other hand, “. . . an epistemically vicious scientist is consistently negligent towards counterevidence or norms of proper inquiry, is culpably incompetent, reports false findings, makes unwarranted scientific inferences, engages in bad methodological practices, irrationally clings to bad theories and models, or in the worst case falsifies or fabricates data” (Uygun-Tunc and Pritchard, 2022). Although the above traits in practicing science are suggestive of some degree of moral responsibility on the part of the scientist, it might not come down to outright moral corruption. In other words, epistemic vices (or the lack of epistemic virtues) could be viewed at the very least as precursors, if not direct drivers of QRP and RM, perhaps with some degree of independence from true moral transgressions or moral failings.

Epistemic virtues and vices of individuals in influencing incidences of RM/QRP

Although it is a general practice to formulate working hypotheses as a starting point in research, scientists should approach a research question or problem with an open mind, both in terms of approaches as well as interpretation of the data/results obtained. In this regard, epistemic vices such as dogmatism, close-mindedness, negligence, or indifference to knowledge would greatly impair knowledge acquisition. Worse, these could lead to QRP or RM, as illustrated in the two fictitious case below.

Consider the case of a graduate student researcher, “A,” embarking on a project assigned to him. A’s investigative vision is constrained by the faculty supervisor’s pet hypothesis, which has some support from immature and preliminary data left behind by another student who has since departed the lab. A worked hard and diligently, but could find no strong support for the hypothesis, nor was he successful in faithfully reproducing the previous preliminary data. Neither A nor his supervisor would contemplate the possibility that the hypothesis was wrong, and the latter continued to suggest that the former were not skillful enough to reproduce previous results, which the supervisor deemed promising. With assessment deadlines nearing and acting out of desperation, A cherry-picked data points from his many experiments to form “presentable” graphical constructs that approximate the previous data and in support of the supervisor’s hypothesis, and reported these. A denies that he had cheated, and asserted that his thoughts were that the supervisor and his hypothesis must be right, but the experiments are just “difficult to do” and the data are “all over the place.” He has thus acted in the best interest of himself and the lab, but the act amounts to data falsification and would be a clear case of RM.

Postdoctoral fellow “B” is tasked to analyze the presence of a particular protein biomarker in human samples using a technique that is dependent on a commercial antibody reagent. The technique appears straightforward, with the antibody and an associated reagent producing an electrophoretic migration signal band indicating a known size of the marker being probed. B thinks he knows what the size of the marker is, as it would be in accordance with its coding DNA size, and the same antibody has been used successfully by others. B ran his first experiments but was puzzled by the appearance a signal band that is lower than the expected size. Thinking that the band must be the signal of the marker he is after, B got down to work on the large amount of samples piling up in his freezer. He quantified the band signals from a cohort of participants in a randomized drug trial and presented these with a marking indicating that the bands migrated at the expected size. These results would then constitute part of the information gather on the efficacy of the drug being tested. In this instance, B has falsified research data, and this act came to light as others found his marker distributions haphazard and illogical. B admitted that he reported the marker size wrongly but argued that the signals he saw and reported are authentic. Unbeknownst to B, the band he observed is actually a false signal because the marker protein is heavily post-translationally modified and would appear as a higher band than the expected size, not lower. Furthermore, the antibody he used was a different lot from previous, which gave a spurious signal rather than the right band.

For individual A in the first case above, the researcher appears to be seeking results of some kind to fulfil academic expectations rather than the truth, and thus appear to be deficient in epistemic virtues, particularly in terms of intellectual courage and integrity. The inability to accept that his experimental results might have falsified the supervisor’s hypothesis is an epistemic failure, and is entirely misaligned with the Popperian (Popper, 1992) notion that hypotheses must be testable and conceivably proven false. A more epistemically virtuous researcher would have not only questioned the validity of the supervisor’s hypothesis, but also the validity of the preliminary results left behind by the previously departed student, even if this means confrontational meetings with the supervisor and perhaps incurring a lot more work on his part. On the part of the supervisor, we see dogmatism at play, and this might have rubbed off on the student, with the latter also exhibiting the trait of closed-mindedness. For individual B in the second case, the researcher appears to allow epistemic vices in the form of negligence and indifference to prevail. That a signal at the wrong size is obtained would have immediately alarmed a more epistemically virtuous researcher that one is looking at a false positive. All attempts should then be made to resolve this size anomaly before any samples are examined.

It is notable that in both cases above, a lack of epistemic virtues and a prevalence of vices appeared to have initiated, led, or otherwise ushered the individuals A and B toward committing RM. There is no indication that either perpetrators started off with an intention to cheat, but being epistemically inept and subsequently under some pressure to deliver, they crossed the ethical line or tumbled down the slippery slope. It might be fair to say that had these researchers be imbued with proper epistemic virtues, their fall into moral disgrace could have been effectively prevented.

Epistemic vices of the institution in promoting RM/QRP

Organizations can also exhibit a deficiency in epistemic virtues and suffer from epistemic vices (Baird and Calvard, 2019; de Bruin, 2013; Meyer, 2023; Meyer and Choo, 2023). Baird and Calvard, for example, outlined four types of epistemic vices in organizations, namely malevolence, insouciance, hubris, and injustice, all of which could lead to unethical conduct (Baird and Calvard, 2019). It is widely known that research culture and governing measures of an institution would make RM/QRP more or less conducive. Epistemic vices on the part of organizations and institutions of research could likewise contribute to RM/QRP. In terms of research output, the institution should ideally act as a gatekeeper for research authenticity and quality. However, constant auditing of research results and manuscripts sent for publication is often deemed unwieldy and undesirable as this might generate an atmosphere of distrust and stifle academic freedom. There is also a tendency of institutions to protect their own reputation and those of faculties with stellar academic performances as well as the revenues (external grants, patents) they bring in. Very often, however, the institution is then left to deal with the aftermath of RM, wasting resources in handling complaints and whistleblowing and suffering from tarnished reputations. I posit that an epistemically virtuous research institution may pursue non-epistemic goals (such as revenues and status, as individuals do), but should always make it such that the epistemic goal of contributing to knowledge discovery is prioritized.

Consider the following case in which Professor “C” is a top researcher in a university who is endowed with multiple large grants and runs a consortium of international research collaborations. Despite C’s stellar performance and status in the research community, there are intermittent murmurs from the ground on irregular research practices in his lab and on suspicious data in his published papers. The university had ignored these warning signs, only to be blindsided by a flurry of allegations of research misconduct and harassment made by discontented students and junior staff. C insisted, despite overwhelming evidence to the contrary and editorial retractions of his papers by journals, that he is unaware of any of the RM incidences, and that he would not have allowed any fraudulent material from his lab to be published. He also did not think that some of his persistent and aggressive pushes for data and crude jokes made during his group’s social gatherings could be construed as harassment. The university realized, albeit in retrospect, that three highly promising junior faculties have all opted to leave the university not long after being drafted into the collaborative program which C leads. ³

In the case above, although C appears to be the one directly at fault, the institution has made it conducive for him to act aberrantly. While it is unclear how the RMs occur, but as the head of the lab, C has to take definitive responsibility. It is not clear if the institution is morally at fault or blameworthy in this case, as most institutions would not have conducted any formal investigation unless there are officially registered complaints. However, by ignoring warning signs, the institution has exhibited the epistemic vice of indifference or insouciance. Although the institution could not and should not be expected to closely meddle with the research affairs of its members, it should be mindful and sensitive to potential employee issues and act promptly. ⁴

Epistemic vices of reviewers, editors and journals in promoting RM/QRP

Academic peer review is a necessary and quality gatekeeping mechanism before scientific results are published. Epistemic vices exhibited by those involved in the peer review/publishing process would also promote RM/QRP. I illustrate this possibility with another hypothetical case below.

A journal receives a manuscript and the editor “D” sent it for a single-blind peer review to two reviewers (labeled as “E1” and “E2”) who responded positively to the invitation. E1 appeared to be skeptical about the paper’s theoretical foundations and the supporting data, cited several instances of anomaly and irregularity in the data and images of the manuscript, and recommended major revisions with additional control experiments. E2, a senior figure in the field of work has, however, made glowing remarks and recommended only cosmetic changes, as the authors’ findings are well aligned with E2’s own ideas and theories. D conveyed the reviewers’ comments to the authors with an editorial disposition of “major revision”. The authors reverted with a revised manuscript after a week with all the recommended cosmetic revisions made, but no further experiments were added. In the second round of review, E1 remained negative, while E2 recommended acceptance. The editor accepted the paper, and the final accepted draft was published online immediately by the journal. However, within the next few days hours, multiple postings have appeared at the online forum PubPeer (Ortega, 2021), noting various irregularities amounting to strong suspicions of data falsification/fabrication in the accepted manuscript, to which the authors offered no reply.

What are the issues here? Did any of the characters behave unethically? Again it appears that moral fault, if any, lies squarely with the authors but the peer review-publication system had in this case made such RMs more conducive then it would otherwise be. It appears that reviewer E2 has probably not scrutinized the data and results of the manuscript as thoroughly as E1, and was uncritical with regards to the data and image irregularities which E1 has duly noted. There is little doubt that finding suitable peer reviewers for a manuscript is now becoming an arduous task. There are some indications that this could be down to reviewer fatigue (Fox et al., 2017; Vesper, 2018). In any case, this difficulty might have influenced the neglect of seeking a third opinion when the opinions of the initial reviewers are polarized. However, to insist on getting another review on board in this case would be more epistemically virtuous on the part of the editor and the journal. Furthermore, when data or image irregularities are highlighted, it would be prudent for the editor/journal to request for the raw data files from the authors for scrutiny (if these are not already supplied). Failing to exercise such due diligence would not exactly be unethical, but would certainly be associated with the epistemic vice of negligence. The stringency of peer review is of paramount importance as it is the final and often the only true gatekeeper of research quality. If this is done inadequately, as in the case of predatory journals, the trustworthiness of the work would be compromised. ⁵

Journals and their publishing houses are business establishments which might be expected to place their priorities on profit and revenues. However, as primary outlets of knowledge it is critical for these to have adequate epistemic virtues. Implementing a stringent peer review for quality control would be an aspect of such virtue. Beyond relying on reviewers, however, journal editorial offices should be equipped with adequate resources to detect data and image irregularities, and with editorial staff trained to perform such tasks. As per the case example above, data or image irregularities mentioned by E1 should have triggered a careful examination of the said manuscript’s submitted materials even if this is not done routinely for all submissions. Striving to ensure the validity and authenticity of data and results would aid the attainment of true epistemic goals, and would be epistemically virtuous.

Improving epistemic virtues and addressing epistemic vices in the research ecosystem

The examples given above point to the possibility that a deficiency in epistemic virtue might be fairly common. A deficient epistemic virtue and epistemic vices could play a precursory if indirect role in promoting incidences of RM/QRP. In other words, a researcher that is not sufficiently determined to place epistemic goals above personal gains and feelings would be susceptible to commit RM, and these aberrant acts could be made conducive by the analogous epistemic virtue deficiencies within the research organization or institution, as well as that of the peer review-publication system. Although modern day scientific endeavors cannot be value free and is influenced by social and economic factors (Elliott, 2017; Longino, 1990; Resnik, 1996; Resnik and Elliott, 2019), science is still fundamentally an epistemic pursuit. In the sciences, knowledge collected through research enable understanding of the natural world. Scientific results and their interpretation should thus at the very least be, epistemically speaking, justified true believes. To be gainfully engaged in research and yet not place epistemic goals above all else would be a fundamental fault that arguably goes beyond moral virtues and convictions.

What then, can we do about deficiencies in epistemic virtues and prevalence of epistemic vices in the scientific ecosystem? Epistemic virtues and vices are character traits that might not be easy to change. For the individual, education endows epistemic capabilities and values. However, the STEM curriculum does not always adequately prepare a young researcher to do research in an epistemically optimal manner. I have previously discussed the importance of infusing elements of philosophy of science into the syllabi of research ethics or RCR education programs (Tang, 2023a). Likewise, such courses may also be a media of illustration for epistemic virtues and vices. An individual should be made to reflect on his or her own shortcomings in terms pursuing epistemic goals, to be encourage to do better and to not let vices get into the way of their research work.

To foster or inculcate epistemic virtues at the individual level, various strategies have been proposed for implementation in higher education (Kuntz and Taylor, 2021). For example, Orona et al. have reported experimental successes in enhancing virtue learning among undergraduates (Orona, 2021; Orona and Pritchard, 2022; Orona et al., 2023). In an issue of Education Psychologist on “Educational responses to a ‘post-truth’ world: Diverse approaches to improving thinking about scientific issues,” the education sector’s responses to epistemic deficiency are elaborately discussed (Barzilai and Chinn, 2020). Lapsley and Chaloner have suggested that “science education is in the business of character education,” as “intellectual character is strongly metacognitive and requires a concept of science identity to provide a motivational force to the work of virtues” (Lapsley and Chaloner, 2020). On the other hand, scholarship in psychology and humanities would reciprocally contribute towards the educational drive in promoting epistemic virtues in the social and natural sciences.

For the institution, beyond regulatory and oversight measures, cultivating a culture of openness, inclusiveness, equity and healthy mutual criticism would promote epistemic virtues and suppress epistemic vices. Giving a researcher too much control and executive power beyond managing the research itself could often allow the vices of one that would affect many others. A research institution must want to be in the know of the work culture and atmosphere in the groups and laboratories, and should have mechanisms of balances and checks in place to mitigate vices. An institution should be aware of bad influences by epistemic vices such as close-mindedness, negligence or indifference that promote a poor research culture. Other than putting in place and mandating trainings in responsible conduct in research and whistleblowing channels, institutions should be mindful about indications and warning signs of wrongful acts, and not ignore these. The institution can, for example, conduct periodical lab visits and audits to better gauge signs of injustice and grievances in the research setting arising from surveys and feedback mechanisms, particularly those from junior researchers. Building channels “to keep an ear on the ground” would be epistemically virtuous in this regard.

RM and QRPs are arguably rooted, in a large part, to the highly competitive, productivity oriented, publish-or-perish academic culture. Other than inadequate institutional oversight (Yeo-Teh and Tang, 2022), publishing pressure appears to be a major promoting factor for RM (DuBois et al., 2013; Paruzel-Czachura et al., 2021). Due to a superficial and wholly inadequate emphasis on publication output on career advancement and other rewards, the culture is such that there is a wrong tendency for scientists to focus on maximizing the number of papers rather than their genuine epistemic values. In appraisal and recruitment exercises, the institution must move away from emphasizing output quantity to one that values quality. It is pleasing to note that there are now multiple examples for such changes, and that the Hong Kong Principles ⁶ recommended assessment of researchers based on terms other than publications.

Journals should want to make the adjudication and dissemination of knowledge their foremost priority beyond impact factors, rankings and profits. This would mean being epistemically virtuous in subjecting all manuscripts to a peer review process that is thorough, stringent and fair, even if this means a longer time for turnover and publishing a smaller number of papers. No effort should be spared in validating and authenticating suspicious data or results, even if this require investments in equipping the editorial office with the necessary software, hardware and human expertise that would aid such authenticating activities. This would be particularly important as we face a potential onslaught of artificial intelligence-generated data, images and manuscripts (Elali and Rachid, 2023; Wang et al., 2022).

Concluding remarks

I have elaborated in the writings above the notion that deficiencies in epistemic virtues and a prevalence of epistemic vices could have an initiating or promoting role in RM/QRP, perhaps more so than a proposed lack in moral virtues. If so, efforts should be made to promote epistemic virtues and to mitigate epistemic vices. These efforts are important, as a perception of epistemic virtue affects how the lay public views and appreciates the researchers and their institutions. Piovarchy and Sisikind have recently expounded on the general concept of epistemic health, “. . . a measure of how well an entity (e.g. person, community, nation) is functioning with regard to various epistemic goods or ideals” (Piovarchy and Siskind, 2023). It is obvious that epistemic health is critical to the research enterprise and all within the research ecosystem would do well in staying epistemically healthy.

The promotion of epistemic health in science should be targeted, as suggested by the discussion above, toward the individual, the institution and the associated publication ecosystem. In this regard, the major challenges would be how epistemic health could be adequately evaluated in these three different groups of targets, and how interventions could be implemented and their effects measured. For the former, separate evaluation matrixes, comprising of relevant and appropriately weighted epistemic virtues and vices, could be designed for the three groups. For the latter, education and training on epistemic virtues should be mandated for the individual researcher, while centralized or community-endorsed rules and best practices should apply to institutions and journals/publishers.

Advances in technology would typically open new horizons in scientific research but could also pose challenges to upholding of epistemic virtues and promote epistemic vices. A pertinent example would be the application of artificial intelligence (AI) in research (Chubb et al., 2022), of which the advent of generative AIs in the form of large language models (LLMs) such as OpenAI’s ChatGPT is but one aspect (Ramponi, 2023). With ChatGPT, manuscripts that are semantically accurate, logically sensible and factually correct to a large extent could be generated very quickly and easily by a machine from short prompts. As journals move to restrict or mandate declaration of their usage (Hosseini et al., 2023), it is clear that ChatGPT and other rapidly advancing generative AI tools shall change the way in which research is done and how papers are written (Mann, 2023), but unfortunately also how RM/QRP could be more conveniently or inadvertently committed (Gu et al., 2022; Tang, 2023b). The importance of epistemic virtues in the use of generative AI in research and publications could not be understated, and epistemic health in research is more important now than ever.