Why should we care? Shepperd and Dalderop (2025) have asked a powerful question in the title of their paper. As scholars, we care that our research is accurate, meaningful, and connected to the larger literature. However, as Shepperd and Dalderop (2025) remind us, we should also care about our participants. We want to minimize harm caused by our research, but we also want the people who take on any risks associated with participating in our study to benefit in some way.
The authors of the current paper have been working closely over the last five years on a series of projects related to research ethics and questionable research practices (QRP s). We do not explicitly research low literacy language users, nor do we couch our understanding of research ethics in an ethics-of-care framework. However, our presentations and workshops often include conversations on the limited connection that ethical review boards have to the daily micro-ethics (Kubanyiova, 2008) involved in applied linguistics research and the need for researchers to consider how their decisions impact all aspects of the research process. In reading and discussing the article by Shepperd and Dalderop (2025), we found ourselves in agreement with much of what the authors wrote. This piece also led us to consider and discuss a multitude of related topics. What follows are our thoughts on three topics that were spurred on by the original article. Our hope is that scholars will consider these topics as they read Shepperd and Dalderop (2025) and then reflect on their own day-to-day research practices.
Applied linguistics has witnessed an ongoing trend towards the principles and practices of open science and open scholarship. For many of us, this includes making our research available to a larger and sometimes non-academic audience (Plonsky, 2024). Open research can involve, for example, publishing outside of paywalls, making data and findings more accessible, and sharing our work in venues designed to reach non-specialists (e.g., Andringa et al., 2024). These practices potentially give more stakeholders access to research. However, making research more open does not necessarily make it more understandable to or useful for a wider audience. Open-access materials are often written with a scholarly audience in mind, meaning that they might be challenging to read for many people. If our studies are purposefully targeting populations of lower language levels, then this challenge is likely to only be more extreme. In short, just making our materials open does not by default make them meaningful or useful to stakeholders or the community members we study. Asking participants to take on risks when they do not appear to be able to utilize the knowledge they help generate might be considered a QRP.
We are not in a position to insist that applied linguists should focus their work on science communication or community engagement activities. Even in an applied field such as ours, ‘basic science’, also referred to as ‘bench research’ has its place. And more theoretically oriented work, for example, can lead to new knowledge and to the development of new ideas that can impact the lives of many people. Instead, we argue that scholars should consider the utility of their research and find ways in which their work can impact the groups that participate in their research (Jenks, 2025). As a field, we might want to consider ways of communicating our science (Wagner & McKee, 2023) or utilizing community-engagement practices. Making our research useful to the groups we are working with is one method of ensuring an ethics-of-care.
Sterling et al. (2016) investigated the type of ethics training that applied linguists received and found that much of the training centered on topics connected to ethical review boards. In most cases, this training was carried out as part of a research methodology course and/or through informal training received through mentorship. However, the efficacy of formal ethics training has been called into question (Antes et al., 2009). With the general increase in publications on research ethics within the field (De Costa et al., 2024; Yaw et al., 2023) it is unclear how reliable the findings from Sterling et al. still are. However, Wood et al. (2024) found that textbooks and syllabi used in research methods courses were still focused largely on issues related to ethical review boards. It is also the case that many applied linguists have admitted to making questionable decisions in their research (Isbell et al., 2022; Larsson et al., 2023). Many publications have remarked on the need for research ethics training to move away from a heavy focus on ethical review board issues, what is often called procedural ethics (Guillemin & Gillam, 2004), into something that resembles the ethical challenges that scholars face on a day to day basis (Kubanyiova, 2008).
Our own work (Yaw et al., 2026) has isolated mentorship as a key area for consideration. We often think of mentorship as taking place between students and an advisor, but that is only one potential area. All scholars in the field are likely to face ethical challenges and require support from time to time. We should certainly consider policies and practices designed for student mentorship, but we can also work to find ways of developing continual education for all scholars at all career stages in the field. We can take the emergence of AI in the last few years as an example. There are many questions about when to use AI, how to report its usage, and more. Even the most experienced scholars among us may be unsure of how to address these concerns. It’s critical, therefore, that we encourage and support each other to improve not only one-on-one mentorship but continual research ethics training for all.
Admissions of convenience sampling are often heard in the same breath as disclaimers and apologies. Some of the major issues with this form of sampling are that it can introduce different forms of bias and can greatly limit generalizability. The general recommendation to avoid convenience sampling is well-founded and one that many research projects should adhere to.
However, as with most QRP s, the use of convenience sampling is neither ‘good’ nor ‘bad.’ There are many cases in which it makes sense to utilize the resources that one has access to. As an example, it is practically impossible to collect data on intact classrooms using some form of randomized sampling. In fact, claims of the threat to internal and external validity posed by a lack of random sampling should be weighed against the enhanced ecological validity found in classroom-based research. We might want to work with specific communities and community members, which can increase one’s ability to collect meaningful data. It might also be possible to generalize these claims to a larger population. In terms of QRP s, the goal in sampling is to be transparent but also understand your own data. Being transparent with how we collected data and making claims that fit our data is likely to be an understandable research decision. Knowingly making claims that exceed the information we can obtain from our data and representing it as generalizable is questionable.
Reading the Shepperd and Dalderop (2025) article forced us to stop and think about our own ethical practices, an act we regularly do in our group meetings. Ensuring that our research matters to the people we collect data from requires us to consider how we publish our results, how we educate scholars within our community, and how we sample from various populations. In starting these conversations, our hope is that other scholars will likewise pause to reflect on their own decisions throughout the research cycle.