This editorial inaugurates a series of three, dedicated to the theme of open science (CA) and its implementation in the scenario of research in Administration. In the first part, we explore the fundamental issues underpinning CA, the challenges that motivate it, such as the reproducibility crisis, and the need to engage advisors and graduate programs in adopting these principles. We call on the scientific community to recognize CA as the foundation for new researchers’ reliability, dissemination, and training. In the following editorials, we move on to the practical dimensions of CA. The second text will discuss the available tools and workflows that an open scientist can incorporate into their research routine. Finally, the third editorial will address the challenge of implementing these practices, proposing a practical template to guide researchers and graduate programs in systematically adopting CA in their projects.

Contemporary science faces profound challenges regarding the reliability of its results and the democratization of knowledge. Multiple fields have experienced the so-called reproducibility crisis, which is the difficulty of reproducing published findings. More than two-thirds of researchers have tried and failed to reproduce other scientists’ experiments, and about half were unable to even produce experiments themselves. Large replication efforts have revealed success rates in psychology and biomedicine, sometimes less than 40% (Baker, 2016). This crisis, widely publicized in magazines and mass media, catalyzed a movement of self-reflection in science. How can confidence in scientific results be restored? Perhaps a consensus has been adopted for CA practices (Munafò et al., 2017). The ‘replication crisis’ made headlines and drove the adoption of CA to improve the reliability of scientific findings.

CA proposes transparency in the stages of research - from planning to dissemination - seeking to make methods, data, and results accessible to academic peers and society. According to United Nations Educational, Scientific and Cultural Organization’s definition (UNESCO, 2022, p. 7), it is

An inclusive construct that combines diverse movements and practices, aiming to make multilingual scientific knowledge openly available, accessible, and reusable for all, to increase scientific collaborations and information sharing for the benefit of science and society, and to open up the processes of creating, evaluating, and communicating scientific knowledge to actors in society beyond the traditional scientific community.

In short, CA means sharing results and building a more rigorous, collaborative, and transparent research ecosystem in which knowledge flows freely and benefits society (UNESCO, 2022).

The reproducibility crisis has exposed structural flaws in conducting and reporting research. Publish or perish, pressure for positive results, and non-transparent methods have accumulated robust evidence. Ioannidis (2005) questioned “Why are most research findings falsified?” - arguing that biases and flexibility of analyses produce false positives. A survey in Nature revealed that 90% of scientists agree that there is at least a ‘mild’ reproducibility crisis, and 70% have already failed to reproduce other people’s results (Baker, 2016). These data shook confidence in the scientific literature and motivated initiatives to correct course. CA emerged as a set of practices to address these problems directly. For example, open sharing of data and code allows others to verify analyses and test secondary hypotheses, reducing inadvertent errors or fraud. Pre-registration of studies and registered reports seeks to mitigate publication and p-hacking biases by documenting hypotheses and analytical plans before data collection. Guidelines such as the TOP Guidelines (Transparency and Openness Promotion), released in 2015, recommend that journals require transparency in materials, data, and pre-registrations. These structural changes aim to put reliability back at the center of the scientific process.

The effects are already noticeable. Studies indicate that the adoption of open practices increases reproducibility. In psychology, methodological reformulation efforts have shown that rigorous practices, such as pre-registration, transparency in procedures, and adequate sample sizes, have raised replication rates of new studies to nearly 90%, a 40-50% jump from what was previously observed (O’Grady, 2023). In other words, CA has demonstrated, in practice, how to improve the soundness of scientific results. This improvement in rigor benefits academia and society, which rely on credible evidence to formulate policies, technological innovations, and health treatments.

In addition to addressing the reproducibility crisis, CA acts as a catalyst for democratizing knowledge. Promoting open access to publications and data reduces informational inequalities between researchers from different countries, academia, and the public (UNESCO, 2022). Open access initiatives, such as the SciELO library in Brazil and policies of funding agencies, have made thousands of articles freely available, expanding their reach and impact. When locked behind paywalls, scientific knowledge limits its potential to generate development and state-of-the-art mapping by generative artificial intelligence platforms. For this reason, more and more journals and institutions adopt the principle that publicly funded research results should be publicly accessible. This movement boosts scientific dissemination: researchers start to dialogue more with the public, whether through open media, institutional repositories, or participation in citizen science projects. Direct exchange with society enriches research, new data and perspectives emerge, and mutual trust is strengthened. Thus, CA fulfills a dual role: it improves the internal quality of science (via rigor and transparency) and expands its external projection (via broad access and social engagement).

In this changing landscape, a new element deserves special attention: the rise of generative artificial intelligence (AGI) in scientific research. Machine learning algorithms now help analyze large volumes of data, find patterns, and even generate hypotheses. However, this tool carries risks if used in an opaque or irreproducible way. Scientists are already concerned that the non-transparent use of AGI is leading to a flood of unreliable or useless studies (Bockting et al., 2023). AI models are often described as ‘black boxes’: if researchers don’t disclose their code, model weights, or training data, others can’t reproduce or validate the results, and it can exacerbate the reproducibility crisis rather than mitigate it. For example, in the race during the COVID-19 pandemic, AGI tools were used to diagnose infections on X-rays. Still, many of these solutions failed to generalize due to a lack of proper validation (Ball, 2023).

CA offers pathways to leverage the benefits of AGI without sacrificing reliability. Open models and open data allow for community scrutiny. Recently, researchers compared open-source versus proprietary AI models, discussing how open, trained, and transparently disclosed models best align with CA principles. The conclusion is optimistic: when applied with ethical rigor, AGI can assist CA, promoting new levels of collaboration and innovation (Ball, 2023). In other words, if we embrace algorithmic transparency - disclosing model architecture, datasets used, and even opening source code - the scientific community will be able to reproduce AGI-guided findings and increase confidence in their use. There are already calls in the literature for the scientific community to establish clear standards for the use of AGI, such as living guidelines that are continuously updated, ensuring that scientists themselves supervise the use of AGI in research (Bockting et al., 2023). This intentional oversight ensures that AGI is employed as a tool for scientific acceleration rather than as an obscure shortcut. Only in this way will the integration of AGI into research reinforce - and not undermine - reproducibility.

In addition to technique, AGI raises ethical dilemmas in scientific communication. Tools like text wizards can generate essays and potentially hallucinate references or facts. Here again, CA suggests solutions: transparency labels indicating whether AGI has aided a text, guidelines for responsible use (several journals already require the declaration of use of AGI in manuscripts), and community involvement in error detection. In short, AGI should not be seen as an antagonist, but as one more reason to strengthen the culture of transparency. Integrated under open principles, it can accelerate discoveries and democratize data analysis; without this, it risks becoming another factor of opacity and crisis.

If CA is the path to more reliable and accessible research, how can we ensure that new generations of scientists embody these values and practices? Here, the role of graduate advisors is highlighted. Advisors don’t just supervise master’s and doctoral projects - they are mentors who shape the scientific posture of their students. Therefore, we argue that advisors should actively incorporate the principles of CA in training researchers at different career levels. It means teaching by example and direct guidance: showing advisees how to plan rigorous and transparent studies, encouraging data sharing and results, and valuing reproduction and open collaboration over unbridled competition.

Empirical evidence already indicates that this transmission of values makes a difference. A recent study published in eLife examined the practices of 211 advisor-advisee pairs in the biomedical field and found a clear pattern: students whose advisors openly shared their data were more likely to also share their data, compared to students whose advisors did not (Haven et al., 2023). In other words, the mentor’s example directly influences the behavior of the future scientist. The same study observed evidence of a similar effect for open-access publications (although in this case, institutional factors also contributed). These findings confirm something intuitive: advisors work as models of conduct. They train researchers aligned with these principles if they practice and encourage CA. Conversely, if a student is never exposed to transparency practices, never participates in an open data initiative, or does not learn how to record a protocol before collecting data, they are unlikely to adopt these practices spontaneously in the future.

Therefore, we call on advisors to explicitly incorporate CA in their research groups and individual orientations. It does not imply diverting the focus from the technical content of the research area, but integrating habits of openness into the scientific routine. It means, for example, discussing data management plans from the beginning of a project, encouraging advisees to publish preprints of their manuscripts (e.g., OSF), or even engaging society (when relevant) in stages of the study. Advisors can also protect students from perverse incentives, such as the pressure for the quantity of publications, by realigning productivity metrics with quality and integrity. Ultimately, it is up to academic leaders to prepare young scientists for a career where collaboration and transparency are advantages, not obstacles. Cultural change in science passes through the graduate classroom and research groups: it is there that the next generation will learn how to do science. They learn from an early age to make it open.

How can advisors apply CA in the training of researchers? Below, we highlight five steps based on these principles to guide the academic advising process. Each step represents a mastery of practices and values that, incorporated into the orientation, cultivate researchers who are better prepared for the current challenges of science.

Graduate programs can transform training practices by implementing these five steps in an integrated way. The student will leave the master’s or doctorate with technical skills in their sub-area and as an agent for disseminating good practices - able to conduct transparent research, collaborate widely, and communicate science beyond academic boundaries. This training investment yields multiplying fruits: each new doctor trained in CA carries forward these principles, impacting colleagues, groups, and institutions wherever they go.

Thus, as a suggestion for researchers to start their research projects or orientations, we propose five stages of CA for graduate programs. These are: (1) transparent planning, (2) data and code sharing, (3) open access and communication, (4) open collaboration, and (5) ethics and integrity. Each axis corresponds to open science practices integrated into the research cycle, offering a practical roadmap for advisors and graduate students to adopt these practices systematically. The idea is that, by following these steps, research groups will make their processes more transparent, collaborative, and reproducible, in line with the FAIR principles (findable, accessible, interoperable, reusable) and with the motto ‘as open as possible, as closed as necessary’ (ensuring maximum openness without hurting legitimate interests, such as privacy or commercial secrecy).

Incorporating CA as a central strategy for scientific development, the dissemination of knowledge, and the training of researchers represents a path of no return - and a promising future. The challenges that motivated this movement, such as the reproducibility crisis and disruptive innovations such as artificial intelligence, are still present and will require constant adaptation. However, the trajectory demonstrates that openness can constructively address such challenges. More than that, CA reimagines scientific work: from an ethos of secretive competitiveness to a culture of transparent collaboration. This new paradigm benefits not only the credibility of science in the eyes of society by making it more reliable and understandable, but also enhances the efficiency of scientific progress through the unrestricted exchange of knowledge and resources.

As a purposeful editorial, we leave a clear message to the scientific community in general, especially to the trainers of new researchers: embrace CA as a daily practice and fundamental value. Each advisor who embraces these ideas helps create a next generation of scientists committed to the quality, integrity, and public utility of research. Every author who publishes openly, every reviewer who acts transparently, and every group who shares their data contribute to an ecosystem in which discoveries accumulate more solidly and knowledge circulates without barriers. There are still obstacles - infrastructure to improve, policies to implement, and, above all, minds to convince. The trend is irreversible; global initiatives, such as the 2021 UNESCO Recommendation on Open Science, and local initiatives, such as the new CAPES evaluation criteria that value open data, show that institutions adhere to this vision.

Each stage - from transparent planning to integrity - acts complementarily, covering the entire research cycle and fostering a more open, collaborative, and responsible academic culture.

CA represents a promising and irreversible path to strengthening the integrity and impact of scientific research. The conceptual discussions and call to action presented in this editorial constitute the first step of further exploration. In upcoming editorials, we will provide practical guidance and concrete tools to assist researchers and graduate programs in transitioning to a more open and transparent workflow. The second text will detail the tools and practices, while the third will propose a structured model to guide the effective implementation of the principles of CA in the daily life of research and the training of new scientists. By embracing CA in all its dimensions - conceptual, practical, and implementation - we will build a more trustworthy, collaborative, and beneficial future of science for all of society.

We conclude by calling on our fellow researchers, young or old: let’s make CA not just a set of guidelines, but the essence of our scientific conduct. By cultivating transparency, collaboration, and sharing in research, we will strengthen the foundations of science, making it more solid internally and more relevant externally. This way, we will fulfill our double social role with excellence: to produce reliable knowledge and disseminate it for the common good. This is the science we want to see flourish in the coming decades.