Abstract
Every year at the Academy of Management meetings, the editors of Group & Organization Management (GOM) meet with the editorial board and friends of the journal to discuss what we are looking for and, just as importantly, what we are not looking for. Usually, this appears as a single slide. The aim of this editorial is to turn that slide into a detailed, written guide for authors and reviewers. We want to clarify what counts as a strong fit and contribution for GOM, how questionable research practices (QRPs) can creep into reviewing just as much as into authorship, how we see the balance between theory and methods at the journal, and how reviewers can appropriately evaluate the different formats we publish—regular manuscripts, GOMusings, GOM Now, and GOMethods—without applying a one-size-fits-all empirical template to everything. Our hope is that this editorial will be equally useful to authors deciding whether a manuscript belongs at GOM and to reviewers thinking about how to write fair, rigorous, and developmental reviews.
The Mission of GOM and What we are Looking for in a GOM Paper
GOM’s mission is to publish work that extends management and organization theory and that has clear implications for practice, with a particular emphasis on groups, teams, multi-level phenomena, and organizational processes. We are especially interested in work that crosses boundaries: across levels of analysis, across national cultures, and across disciplinary perspectives. In line with this boundary-spanning mission, we particularly welcome submissions that draw on underrepresented contexts and voices, including work conducted in Africa, South and Southeast Asia, Latin America, and other regions that have historically been less visible in mainstream management journals. Such work is especially valuable when authors explain how their context helps us see existing theories differently or develop new ones, rather than treating context only as a backdrop. We publish both conceptual and empirical papers, but in both cases, we look for a clear value-added framework, strong argumentation, and a sense of where the field can go next. We also see ourselves as a developmental journal. We care deeply about rigor, but we are also genuinely committed to helping good ideas become excellent papers, and we have developed processes—such as an optional methods review step for some quantitative manuscripts—to support that goal. Importantly, as a specialized journal, we are not trying to be a clone of any other outlet in the field. Our expectations for theory, contribution, and methods are different, but not necessarily higher or lower. In successful GOM papers, theory and methods are in conversation with each other, the focus is on phenomena that matter for groups and organizations, and there is a conscious effort to speak both to scholars and to practitioners who care about how people work together.
A strong traditional GOM paper 1 , whether conceptual or empirical, rests on four pillars: (1) a clear and non-trivial phenomenon and question, (2) a distinctive theoretical contribution, (3) rigor in argumentation and/or methodology, as appropriate to the phenomenon/question, and (4) implications that genuinely matter for how we understand or manage groups and organizations. First, we look for manuscripts that are empirically intriguing, address a previously unresolved theoretical puzzle, yield meaningful practical consequences, or ideally deliver all three. We are especially drawn to questions that are framed in terms of groups, teams, collectives, and organizational processes and outcomes, not only individual attitudes or behaviors in isolation. Moreover, if authors cannot state their core question in one or two sentences without heavy jargon, the work is usually not ready for GOM.
Second, we expect authors to be explicit about what is genuinely new. This might include, but is not limited to, a theoretical mechanism that has not been articulated before, an integration across previously separate literatures, a reconceptualization of an established construct, a novel boundary condition, or a new way to theorize multi-level dynamics. We encourage authors to ask themselves who would need to rethink what after reading the paper. Equally important, we ask authors to spell out why this new mechanism, boundary condition, or integration matters—both for the scholarly conversation and for managers or organizational decision-makers: What can be understood, anticipated, or done differently once this addition is taken seriously? If no one would have to revise any prior assumptions or reconsider how to approach management problems, the contribution is probably too modest for the journal.
Third, we value methodological sophistication, but we do not equate “more complicated” with “better.” The core question should drive the design and analysis, not the other way around. For empirical work, we look for designs that align with the claims authors want to make. If a paper makes strong temporal or causal claims, we expect longitudinal, experimental, or quasi-experimental evidence. If a paper makes multi-level claims, we expect that the proposed levels are actually modeled and that there is a sufficient number of units at each level, and that all effects are argued and interpreted properly, according to the corresponding level at which the effects were statistically observed/found (see Lemoine et al., 2025). Across methods, we value transparent reporting of sampling, measurement, analysis decisions, and limitations. We also expect realistic claims: strong internal validity where the design permits it, and appropriately qualified conclusions where it does not. Moreover, we appreciate authors who are willing to show what they did “under the hood”—for example, in online appendices—without burying the main story under technical detail. What we are less enthusiastic about are “model-of-the-month” approaches where theory is thin but estimation is intricate, underpowered studies dressed up with complex models, or cosmetic fixes to model fit (for instance, untheorized parceling or indiscriminate removal of items and paths) that are not grounded in a substantive rationale.
Fourth, an important theme in our recent editorials (e.g., Griep, 2022a, 2022b; Zagenczyk, 2021) has been that GOM has its own expectations for how theory and methods should interact. We do not see “theory” as something that must always take a very specific form, nor do we see “rigor” as synonymous with a narrowly defined set of methods. Instead, we emphasize balanced contributions in which theory and methods are mutually reinforcing. At the same time, we caution against a “more is more” approach to theorizing. We increasingly see manuscripts that stack multiple theories, models, and long lists of citations into a single paper in the hope of strengthening the argument. Just as methodological sophistication does not mean that “more complicated” is necessarily “better,” sprawling literature reviews and multi-framework rationales do not automatically improve a manuscript. Theory matters, but when too many theories and models are emphasized at once, the central logic of the paper is often obscured rather than clarified—especially when the submission is not a conceptual piece or explicitly seeking to reconcile competing perspectives. For empirical manuscripts in particular, we encourage authors to prioritize a clear theoretical throughline over theoretical accumulation: additional lenses should be brought in only when they sharpen the core argument, not when they turn the front end into a catalogue of possible explanations.
For conceptual work, this means building integrative frameworks, introducing new mechanisms or boundary conditions, or offering compelling future research agendas, rather than merely summarizing existing literature. For empirical work, it means that methods are chosen because they are well-suited to the question at hand and because they enable a meaningful theoretical contribution, not because they look impressive on a methods checklist. We also embrace pluralism in approach: qualitative, quantitative, mixed-methods, simulation-based, and meta-analytic work are all welcome, provided they are executed rigorously and serve a clear theoretical purpose. For qualitative research, in particular (more specific information on this follows), rigor includes a transparent account of the sampling strategy, data generation (e.g., interviews, observations, documents), and analytic approach (e.g., grounded theory, thematic analysis, comparative case analysis), as well as clear links between the evidence presented and the theoretical claims. We are less enthusiastic about qualitative papers that provide only sparse methodological detail, treat a small number of quotes as anecdotal illustration rather than systematic analysis, or use qualitative material solely as decoration for a largely pre-determined model. Reviewers should evaluate qualitative submissions on the depth, coherence, and trustworthiness of their interpretations rather than on conformity to quantitative templates. Moreover, GOM sits at the intersection of theory and practice, so we care about implications. We want papers that offer actionable insights for leaders, teams, HR professionals, or policymakers, and we want those implications to be grounded in the arguments and evidence the paper presents. We also expect suggestions for future research that go beyond generic injunctions to study a construct “in other contexts” or “with other methods.” Similarly, future research sections that list every possible moderator, mediator, and new context without a clear logic are not very helpful.
Finally, to address the concern about “overclaiming” survey responses as behavior, we encourage authors (and reviewers alike) to (1) avoid labeling self-reports as behavior, (2) explicitly qualify what items capture (perceptions, intentions, or self-reported frequencies), and (3) align claims with the research design by reserving behavioral language for indicators with observable/administrative traces (e.g., logs, records) or direct observation. Concretely, throughout the manuscript we encourage authors to replace phrases like “we measure behavior” with “we measure self-reported [construct]” and temper all causal/process phrasing accordingly; we also encourage adding a limitations note committing to triangulation in future work (e.g., passive traces, experiments, field audits) and—where feasible—report simple convergent validity checks between self-reports and any available behavioral proxies. This change reflects a broader literature showing that self-reports and behavioral measures often correlate only modestly because they entail different response processes and reliabilities, so behavioral claims should not be inferred from surveys alone, and survey anchors/design choices can systematically bias frequency reports (Banks et al., 2023). This calibration is also consistent with GOM’s editorial guidance to match the strength of claims to the strength of evidence and to be transparent about design limits and measurement choices.
Special Attention to Qualitative Research at GOM
GOM embraces qualitative research as a central pillar of our scholarly community. In the past years, we have published impactful qualitative articles to advance theorizing at the individual, team, and organizational levels (see our “riddle me this” editorials for examples of some of the stellar qualitative work we published over the years; Gardner, 2015, 2016, 2017, Gardner, 2018, 2019, 2020, Griep, 2022a, 2022b, 2023, 2024, 2025). These contributions have drawn on a range of methods, including ethnographies, case studies, grounded theory, discourse analysis, and narrative approaches. In our editorial team, six Associate Editors—and a substantial proportion of our Editorial Review Board—bring expertise in qualitative methodologies. We welcome work that offers novel theoretical insights through empirical depth, methodological coherence, and contextual sensitivity.
First, we expect authors to make an explicit “fit-for-purpose” argument for why a qualitative approach is the most appropriate way to address their research question. This begins with clearly naming the theoretical puzzle the study seeks to resolve (i.e., what is not yet explained or understood in the literature) and then explaining why the specific qualitative approach adopted (e.g., ethnography, case study, grounded theory, discourse analysis, qualitative longitudinal design) is best suited to illuminate that puzzle. In other words, the method choice should be justified as enabling access to the kinds of mechanisms, meanings, practices, routines, or processes that cannot be credibly captured through a purely variable-centered design. Relatedly, we want to emphasize that qualitative research is fundamentally about in-depth analysis of a phenomenon and about theory building. Strong qualitative papers do not primarily catalogue what respondents said or did, nor do they present loosely connected lists of factors assumed to influence an outcome variable. Instead, they aim to explain the “arrows” in organizational life—how and why patterns unfold—by theorizing the practices, routines, and processes that generate observed outcomes, rather than reproducing box-and-arrow models with qualitative quotes attached. Moreover, as several qualitative scholars have noted, the challenge is not to describe the entire river of qualitative data, but to identify and explain the “golden nugget” within it. In this spirit, qualitative papers should move beyond showing “what this is a case of” to clarifying “what this is a theoretically interesting case of.” We therefore encourage clarity and depth over exhaustive enumeration and clustering—an issue that can be especially prevalent in qualitative team-level research, where authors may default to familiar IMOI-style framings rather than articulating a process theory that explains how team dynamics unfold over time.
In general, we expect qualitative work at GOM to demonstrate both rigor and relevance. By rigor we do not mean a performative alignment with popular frameworks or adherence to procedural checklists, but a transparent, well-argued, and documented account of how data were collected, analyzed, and interpreted. We are increasingly observing authors citing established approaches—such as the Gioia method (Gioia et al., 2013) or Braun and Clarke’s (2006) reflexive thematic analysis—without adequately engaging with their epistemological commitments or applying the frameworks as intended. For instance, some studies claim to use the Gioia method without providing a data structure or process model, or showing iteration between data and theory. Others cite reflexive thematic analysis but omit any notions of reflexivity and positionality (do check the “Reflexive Thematic Analysis Bingo” from Braun and Clarke for a humorous overview of misapplications). These misapplications are a strong signal to AEs that something is off. To overcome such issues, we demand transparency, including the use of data structures, coding tables (anonymized if necessary), visual models of the data collection and analysis process, and appropriate attention to saturation, triangulation, and disconfirming evidence where relevant. We encourage a “show and tell” principle for the write-up of qualitative findings, where claims are illustrated with ample excerpts from interviews, observations, or texts. However, we also expect theorization of the findings and not just a description of events. A useful litmus test is whether the analysis helps readers understand why the phenomenon works the way it does (mechanisms and process), not merely what themes appeared in the dataset. If the findings section reads like an inventory of categories, we encourage authors to tighten the analytic focus: foreground the central mechanism(s), show how they connect, and make the theoretical “payoff” explicit so readers can see the contribution beyond the immediate empirical setting. We are well aware—and many of our Associate Editors share this frustration—that striking the right balance between analytic insight and rich description can feel like walking a tightrope. Too much narrative, and reviewers ask, “Where is the theory?” Too much abstraction, and someone inevitably says, “But give us details—what actually happened?” At GOM, we take a developmental stance in helping authors navigate this tension. We welcome submissions that are still finding that balance and are committed to working with authors to ensure that their findings are both empirically grounded and theoretically meaningful.
Additionally, we encourage authors to use online supplements to share materials such as interview guides, coding frameworks, and extended data excerpts. As Aguinis and Solarino (2019) showed in their analysis of interview-based studies published in the Strategic Management Journal, a majority of papers lacked sufficient methodological transparency to enable even conceptual replication. At GOM, we aim to foster a standard of clarity and documentation that supports trust in qualitative research. We also acknowledge that qualitative research is not always replicable in a narrow procedural sense, particularly when grounded in interpretivist or constructivist paradigms. Nonetheless, we value research that generates theoretical insights with implications beyond the immediate context of a single study. Authors should clearly articulate how their findings contribute to broader conversations, whether through identifying mechanisms, theorizing boundary conditions, comparing across embedded cases, or building mid-range theoretical models.
What we are not Looking for in a GOM Paper
Just as important as what we are looking for is what we are not looking for. In our annual editorial board meetings, we share a slide with a list of “red flags” that quickly push a manuscript toward a desk rejection. Importantly, these red flags are not just about how a manuscript is framed or presented; instead, these are fundamental issues where the developmental approach at GOM would fall short. Here we unpack them in more detail so that authors and reviewers understand both the patterns and the reasons behind them.
Literature Reviews That Only Summarize, or “Bibliometric” Catalogues
We value conceptual and review articles at GOM, but we are not looking for literature reviews that simply summarize what has been done, categorize prior work into boxes, and end with a list of generic “future research directions.” Similarly, we are not enthusiastic about bibliometric analyses that primarily present clusters, co-citation networks, or keyword maps without adding interpretive depth or theoretical insight. A strong review or conceptual piece should help readers see the field differently: It should surface tensions, contradictions, or blind spots; propose integrative frameworks; or offer new ways of theorizing phenomena. Reviews that merely tell us how many papers have looked at construct X in setting Y and then suggest “studying X in other settings” do not move the conversation forward. Bibliometric tools can be useful, but they should serve a larger conceptual purpose. If the review would still look essentially the same if we replaced the focal topic with another–that is, if the structure is generic–then it is probably not yet at the level we seek for publication.
“Teams” Papers That are Not Actually About Teams
We also see many manuscripts that are framed as team or group research but do not, in fact, study teams in a meaningful way. Common issues include: collecting data from a single person and then treating that person’s responses as representing the “team,” failing to obtain sufficient within-team agreement to justify aggregation, or working with a very small number of teams (e.g., 10–20) while making broad claims about team-level phenomena (with the exception of studying complex or hard to study teams, see Maynard et al., 2025). When we talk about “team” or “group” papers at GOM, we expect the unit of analysis to be taken seriously (and be analyzed at the team level or with multilevel modeling approaches). That almost always means data from multiple members of each team, careful justification for aggregation of individual-level data to the team level, and a sufficient number of data points at the team level to support any meaningful statistical inference at the team level. For quantitative papers, this means having a sufficient number of teams—very roughly, on the order of 50 or more—to support meaningful statistical inference at the team level; following established guidelines for data aggregation (Chan, 1998, 2019); and clearly justifying analytic choices with respect to the team constructs under study. For qualitative research, this means triangulating data (e.g., observations, interviews, written accounts, or archival materials) from multiple team members and offering a critical discussion of what convergence and divergence in team members’ behaviors or perspectives imply for the team phenomenon under consideration. Papers that only have one respondent per team (often the leader), or that rest on a small handful of collectives, should be framed as individual-level or case-based work rather than as team research. When a manuscript claims to be “about teams” but treats the team as a rhetorical label rather than an empirically grounded unit, it raises immediate concerns about fit with GOM.
Overreliance on limited Student Samples and Disallowed Online Panels
Because methods should be aligned with the research question, we are increasingly cautious about participant samples. We recognize that student samples and online panels can sometimes be appropriate, particularly for theory-testing at early stages, for designs where the phenomena are not context-specific, or for experimental manipulations derived from prior field study insights. However, we are wary of manuscripts that rely exclusively on (undergraduate) student samples for questions that clearly concern working adults, organizational members, or leaders, and then make broad claims about organizational behavior. We are even more concerned about exclusive reliance on certain crowdsourcing platforms that have raised serious data-quality issues. By this we mean unmanaged or lightly managed crowdsourcing marketplaces in which participants self-select into studies in exchange for small payments (e.g., MTurk, TurkPrime). These sources raise well-documented concerns about data quality (e.g., inattention, bots, misrepresentation, non-naïve repeat participants). Therefore, when these samples are integrated into a research design, researchers are expected to incorporate appropriate screening mechanisms, attention checks, etc. (see Aguinis et al., 2021 for recommendations). We also note that such crowdsourcing marketplaces may also raise questions about ethical issues such as compensation and working conditions, but these issues are nuanced and complex. We encourage researchers to strongly consider the ethical implications of their participant recruitment and compensation strategies, but except in the case of egregious ethical violations, we will defer to local university ethical boards (e.g., Institutional Review Boards) in terms of evaluating the ethicality of research processes. GOM recognizes that overly restrictive sampling policies may have the unintended consequences of excluding researchers from lower-resourced institutions, thus we raise these ethical concerns cautiously. In contrast, professionally managed survey panels (e.g., panels provided by survey vendors that recruit and maintain respondents and implement quality checks) can be appropriate in some cases, provided that authors clearly describe the sampling frame, screening procedures, and response quality checks and that the panel meaningfully matches the theoretical population of interest (see Peer et al., 2022). When crowdsourcing samples are used at all, they should be clearly justified, carefully screened, and complemented by other data sources. We encourage authors to think carefully about whether their sampling strategy matches their theoretical population, and we ask reviewers to attend closely to whether the sample provides a credible basis for the claims being made.
Mediation or “Temporal” Models on Cross-Sectional Data
Another red flag is the use of complex mediation for ostensibly “temporal” models when the underlying data are cross-sectional. Structural equation models, path models, and/or “illustrated models” with multiple mediators, moderated mediation, or “lagged” arrows between constructs can give the impression of temporal sequencing and causal inference, but when all measures are taken at a single point in time from the same source, there are hard limits to what we can reasonably conclude. At GOM, we are skeptical of strong causal language in such designs, and we are particularly wary when complex temporal or process language (“over time,” “leads to,” “drives”) is used without corresponding temporal structure in the data. Furthermore, when authors implement a lag in the data collection process, there should be theoretical significance to the lag (e.g., Dormann et al., 2020; Dormann & van de Ven, 2014; Griep et al., 2021). In other words, the lag should be justified by the amount of time that one expects that it would take for an effect to emerge. Whether an appropriate lag is one day, one week, or one year is contingent upon the research question. If the research question focuses on an affective response, one day may be an appropriate lag, but if the research question involves complex psychological processes that emerge over more extended periods of time, a lag of a month or more may be justified. This does not mean we will never consider cross-sectional studies; it does mean that we expect authors to align their claims with their design. Mediation in cross-sectional data may be used as a way to test pattern-consistency with a theoretical mechanism, but it cannot establish temporal precedence or rule out alternative explanations. We also recognize that temporal separation between predictors and outcomes can, when thoughtfully implemented, be a useful design choice to reduce common method variance. However, temporal separation for this purpose alone does not turn a study into a longitudinal design, nor does it, by itself, justify strong causal language about processes unfolding over time.
“Redundant” Papers: Same Model, New Construct, no Real Rationale
A recurring pattern is what we might call “model recycling”: manuscripts that take a familiar model (e.g., a standard mediation or moderation framework), insert a new focal construct, and then declare a contribution simply because that specific variable combination has not been tested before. Often, the theoretical section in these papers consists of mapping the new construct onto existing boxes in the model without asking whether the model itself needs revisiting or whether the new construct offers any genuinely new insight. “No one has yet examined X as a mediator between Y and Z” is not, on its own, a compelling reason to publish a paper. At GOM, we are not interested in incremental slot-filling. We want to see a clear rationale for why a new construct is needed, what prior models cannot explain without it, and how adding it changes how we understand the underlying phenomenon. If the only novelty is that a familiar model has been reproduced with a slightly different label on one box, the paper is unlikely to be a good fit.
Old Wine in Ever-More-Specialized Bottles
Relatedly, we often see papers that test very well-established models in ever more narrow or idiosyncratic contexts without a compelling reason for doing so. The logic is usually some version of: “Model A has been validated many times; however, it has not yet been examined among [very specific subgroup in a very specific location], so we investigate whether the same relationships hold.” Unless there is a theoretically grounded reason why this specific context should challenge, refine, or overturn what we already know, such exercises add little to the literature. The problem is not the context itself—we are very open to diverse, under-studied settings—but the lack of a theoretical argument for why the context matters for the theory. Simply showing that a robust model “also works” in a new niche is not enough. We ask authors to resist the temptation to write “does it also work here?” papers, and instead to ask “what about this context helps us see something about the model or the phenomenon that we could not see before?” Without that, the paper risks being “old wine in a slightly different bottle.” When authors test established models in new contexts, we encourage them to make context an explicit part of the research model—for example by theorizing contextual features as boundary conditions, moderators, or mechanisms—rather than treating the setting as a neutral backdrop. In well-motivated work, the context is not just where the study takes place, but a theoretically meaningful element that helps refine, qualify, or challenge what we thought we knew. Indeed, while we appreciate studies in distinctive settings, a distinctive setting alone does not constitute a contribution. Authors need to show clearly how the context helps extend, refine, or challenge existing theory, rather than simply re-labeling a familiar model in a new locale. A cool context is welcome; a cool context plus a clearly articulated theoretical advance is what we are looking for.
Papers that Fail to Adequately Explain Methodological Choices
Even when a question is interesting and the design appears roughly appropriate, we sometimes receive manuscripts that do not adequately explain how the research was done. Examples include vague descriptions of sampling (“we surveyed employees from several organizations” without specifying how they were recruited, response rates, or inclusion criteria), limited detail on measures (no item examples, unclear sources, or ad hoc scales introduced with minimal justification), and opaque analytic procedures (e.g., “we used SEM” with no information on estimation methods, treatment of missing data, or model diagnostics). In some cases, we find major modeling decisions—such as parceling items, dropping factors, or aggregating to higher levels—mentioned in passing but not explained or justified. Similarly, for qualitative studies, common shortcomings include failing to specify the epistemological stance or methodological approach and omitting the interview or focus group questions and providing only minimal detail on data collection procedures. Another recurring issue is insufficient information on the coding process (e.g., the number of coders, how disagreements were resolved, or whether reliability or reflexivity checks were conducted). Finally, when studies claim to develop models or theory, authors often do not clearly describe how codes were generated, refined, and integrated into the final framework, making the analytic process difficult to follow. For GOM, clarity and transparency are non-negotiable. Reviewers and readers must be able to understand what was done, why it was done that way, and what the limitations are. Papers that gloss over these issues, or treat methods as a black box, make it difficult for us to assess the credibility of the findings and are therefore poor candidates for publication. When methods are particularly vague or, worse yet, misleading, a manuscript may not be suitable for peer review as reviewers must have the ability to reasonably evaluate methodology.
QRPs and Transparency
Recently, considerable attention has been given to authors’ engagement in questionable research practices (QRPs)—p-hacking, HARKing, selective reporting, overreliance on low-quality online panels, and so forth. We share these concerns and expect authors to avoid practices such as hypothesizing after the results are known while presenting those hypotheses as a priori, rerunning analyses with many small tweaks until p-values cross arbitrary thresholds without transparent reporting, suppressing null or “messy” results that may be theoretically important, or leaning heavily on data sources known to have serious quality problems without appropriate screening and validation. We do not insist that authors pre-register every study, but we do insist on an honest narrative: Authors should state clearly what was planned and what was exploratory, report robustness checks and specification decisions transparently, and resist the temptation to retro-fit theory to any pattern that happens to emerge in the data. We encourage authors to adopt open science practices (OSPs) to the extent possible by sharing data, analytic code, research materials, and statistical software output when plausible (Castille et al., 2022). The Open Science Framework (osf.io) can be a helpful resource for engaging in OSPs. For guidance on maintaining anonymity while engaging in OSPs, see Obenauer (2025).
Taken together, these patterns—purely descriptive reviews, pseudo-team designs, problematic sampling practices, cross-sectional mediation framed as causal process, “old wine” model recycling without theoretical rationale, and opaque methods and QRPs—are what we mean when we say, in shorthand, “this is not what we are looking for.” Our aim in spelling them out is not to discourage creativity, but to redirect effort toward work that genuinely advances how we understand and study groups and organizations.
The Old Wine That We Are Looking For
While this field is often quick to criticize old wine in new bottles, and we have also done so above, any wine connoisseur knows that there are some high-quality wines that improve with age. High-quality replication research contributes to the rigor of our field and the credibility of management research (Köhler & Cortina, 2021). To the extent that GOM claims to be a steward of rigorous scholarship, we must embrace replication. While management research has taken steps to embrace replication (e.g., Carsten et al., 2023; Kraimer, 2023; Selmer et al., 2023), such steps are unlikely to have longstanding success unless there is an ecosystem that supports replication research. GOM is happy to formally confirm that we consider ourselves part of that ecosystem. While we do not necessarily expect every replication manuscript to make the same theoretical contribution as a traditional empirical publication, we do expect them to be strongly motivated. For example, authors of replication papers should clearly articulate why a previous study warrants replication. Authors of replication studies are encouraged to preregister their research. Replications that are not well-motivated, lack methodological transparency, include unjustified deviations from the original study’s methodology, or are otherwise poorly designed are unlikely to be of interest to GOM. Replication studies should have a single focal study for which they base their methodology. For guidance on designing a replication study, please refer to Obenauer (2024).
Reviewing a GOM Paper
At the core of all good reviews is accuracy. When a review contains inaccurate feedback, authors are left with what feels like an impossible challenge. This challenge is likely to occur when reviewers do not substantiate comments with appropriate citations and examples from the manuscript. For instance, imagine a review that says, “There are a lot of control variables in the model. I worry that bad controls may be biasing estimates.” Without proper citations, the author(s) may not clearly understand the reviewer’s concerns. Even if the reviewer cites relevant literature (e.g., Cinelli et al., 2024; Mändli & Rönkkö, 2023), if they do not state which control variables present a concern and clarify why they present a concern, the author is not positioned to respond accurately and efficiently to the critique. As a result, the author may inappropriately adjust their model or reject a valid reviewer suggestion. While we do not wish for reviews to include a laundry list of items that resembles a stream of consciousness ramble, when synthesizing concerns, it is helpful if critiques are justified with supporting literature and examples from the manuscript.
Similarly, when suggesting additional analyses, it is helpful if the justification is clear. For example, omitted variable bias poses a serious threat to causal inferences. However, a variable must be correlated with both the explanatory variable and the dependent variable for its omission to bias results (Antonakis et al., 2010; Busenbark et al., 2022). Unfortunately, sometimes reviewers ask authors to include additional control variables or discuss their omission as a limitation without establishing that this condition has been met. For example, in gender research, reviewers sometimes comment that variables such as ability, performance, and education may influence outcomes (e.g., compensation, evaluations). This may be true, but there is little, if any, evidence that men and women systematically differ on these variables (e.g., Castilla, 2005; Castilla & Benard, 2010). Therefore, asking authors to collect these controls when they are unavailable or discuss their omission as a serious limitation can be problematic, unless the reviewer has provided support for the argument that these variables meet the criteria for omitted variable bias to be present. While this is just one example, it is representative of a larger issue in reviewing: asking authors for everything but the kitchen sink is not particularly helpful (for the author or the Associate Editor) unless requests are clearly justified.
Building off this, Allen and colleagues (2025) discussed the tendency for reviewers to overlook the expertise of authors. In particular, they discussed an example where a subject matter expert felt as though they were treated like a novice. This speaks to the difficult needle that reviewers must thread. We want you to share your expertise with authors, but we also want you to be receptive to their expertise. At times, authors may reject your suggestions and, when this happens, it may be frustrating. We ask reviewers though, rather than immediately digging in and arguing for a specific point, to please consider if the suggested approach is plausible. While there will be times where the proper way of doing something is clear (e.g., do not treat categorical variables as continuous variables in regression), there will be other times where the right approach to investigating a problem may not be as clear (e.g., which theoretical lens is ideal).
We simply ask reviewers to remember that they are not ghost authors and that the author should retain ownership of the manuscript as long as retaining ownership does not compromise validity or scientific integrity. Please refrain from requesting additional studies that are not necessary for the core theoretical mechanism, particularly when they would be difficult or unethical to obtain, simply because other journals often publish multi-study packages. Similarly, please remember that our personal methodological preferences are not necessarily universal standards. Treating preferences as standards can result in requiring an unnecessarily complex modeling approach for simple questions, insisting on specific reliability thresholds without regard to construct breadth, or mandating the use of one estimation method over another without a clear justification.
Less often discussed, but equally important, are the emergence of QRPs in reviewing. Reviewers can unintentionally push authors toward QRPs. This happens when reviewers implicitly encourage HARKing—for instance, by telling authors to write theory to match post hoc exploratory results and to present those as formal hypotheses. It happens when reviewers demand a significance chase, asking authors to keep trying alternative analyses until a particular effect becomes statistically significant. Building off this, reviewers may unintentionally imply that authors should engage in QRPs when their criticisms focus on the significance of findings (e.g., “this may be the reason that you failed to find significant results”) rather than the scientific merit of methodology (Obenauer & Griep, forthcoming). Finally, QRPs in reviewing include self-serving citation requests, especially when reviewers press for extensive citation of their own work that is not substantively central to the argument. In all of these cases, reviewers risk becoming co-designers of questionable practices and contributing to a research record that is less credible than it appears.
What we expect from reviewers instead is a focus on conceptual clarity, design appropriateness, and transparent reporting rather than on whether every coefficient is significant. We encourage reviewers to welcome exploratory analyses, provided they are labeled as such, and to support authors in clearly distinguishing between confirmatory and exploratory elements. We ask reviewers to emphasize effect sizes, confidence intervals, and theoretical meaning over arbitrary p-value thresholds. Requests for additional data or new studies should be made sparingly and only when they genuinely strengthen the core argument and are realistically feasible.
When suggesting additional citations, reviewers should focus on work that meaningfully advances the argument and be sensitive to potential conflicts of interest. These suggestions can include the reviewer’s own publications, but when suggesting that authors cite this research or that of close colleagues, please be sensitive both to potential conflicts of interest and to reviewer anonymity. In some subfields, especially where only a small number of scholars work on a specific topic, explicitly flagging which citations are yours may effectively reveal your identity. In such cases, an alternative is to share any relevant self-citations in a confidential note to the Associate Editor, who can decide whether and how to pass them on to the authors. When you do list citations directly in the review, they should be offered as options rather than requirements, and you might use language such as: “These papers may be a useful starting point; please do not feel compelled to cite them if you identify more appropriate references.” For qualitative submissions in particular, we also ask reviewers to refrain from imposing paradigmatic assumptions that are not aligned with the study’s design. Before, for example, requesting intercoder reliability statistics or quantification of codes, reviewers should first ask whether such expectations are compatible with the researcher’s stated paradigm and method. A constructivist ethnography and a post-positivist multiple-case study entail different forms of justification and evidence. We expect reviewers to assess credibility, transparency, and contribution to theory without privileging one methodological tradition over another. Our goal is to maintain a genuinely pluralistic space for qualitative research at GOM.
Reviewers are often concerned with ethical issues involved in data collection, and rightfully so. When such issues arise, we ask reviewers to inquire about local ethical review (e.g., IRB approval). If a research process was approved by the local ethics review committee, we ask reviewers to defer to the committee’s decision, even if the reviewer would not personally support such a process. However, when there are concerns that a lack of clarification around data collection processes could encourage future unethical behavior, reviewers can feel comfortable asking authors to describe the safeguards that were put in place to protect participants.
Finally, if you are reading this editorial as an author or interested scholar rather than as a current reviewer, we would like to explicitly invite you to consider joining our reviewer pool. GOM depends on a broad and diverse community of reviewers to sustain rigorous, developmental evaluation of submissions. Serving on our reviewer board typically involves reviewing a small number of manuscripts per year (for most people, around three to four), and we are especially keen to include colleagues from a wide range of regions, methodological backgrounds, and topical expertise. If you are interested in serving as a reviewer for GOM, please feel free to contact the editorial office or one of the editors with a brief note about your areas of expertise and a copy of your CV.
Different GOM Formats Equal Different Writing and Reviewing Strategies
One area where our expectations are often misunderstood is the variety of formats we publish. A recurring problem in the review process is that reviewers apply the template for a full-length empirical article to every submission, including those in our special channels. It is worth clarifying the distinct purposes of each format and what reviewers should look for.
A Note on the Usage of Generative AI
Because many authors (and reviewers) now use generative AI tools in their writing process, it is important to clarify our expectations in that domain. We recognize that AI-based tools can be helpful for correcting grammar, improving clarity, and polishing the flow of writing, and we consider this similar to using advanced spelling or style checkers. However, AI tools may not be used to generate content (e.g., ideas, arguments, theory development, literature reviews, or text sections), to write the paper, or to conduct or interpret any form of data analysis. In other words, using AI to “draft” paragraphs, create results, or perform analyses and then merely reviewing the output afterwards is not acceptable. Authors (and reviewers) remain fully and solely responsible for the content of their manuscripts. AI tools cannot be listed as authors, nor can they be treated as co-analysts or co-writers. When AI has been used beyond basic grammar and style checking (for example, to rephrase existing author-written text for readability), we ask authors to include a short disclosure statement indicating which tool was used and for what limited purpose, and affirming that the authors reviewed and edited the output. Routine use of spelling or grammar checkers does not require disclosure. For reviewers, the mere fact that such limited AI support was used is not, in itself, a reason to down-rate a manuscript, as long as the work otherwise meets our standards for originality, rigor, and integrity.
For reviewers, the use of AI is a bit more complex. While authors retain ownership over their intellectual property and can share it at their discretion, reviewers agree to maintain confidentiality when reviewing a manuscript. Consequently, just as a reviewer could not post a manuscript they are reviewing online, they cannot upload a manuscript to an AI tool that will incorporate the manuscript into its training data or otherwise share the intellectual property of the author. If reviewers wish to utilize AI in the review process, they should (1) use a tool that assures confidentiality, (2) disclose what tool was used, and (3) explain how they ensured that the contents of the manuscript will not be shared. Reviewers are responsible for ensuring that the content of their review is accurate and reflects their personal assessment. In particular, reviewers should not rely on generative AI tools to generate the substantive content of their reviews or to produce reference lists. Such tools can “hallucinate” claims and citations, and any arguments or references included in a review must be ones the reviewer has personally checked and stands behind. Our strong preference is that AI, if used at all, be limited to minor editorial assistance (e.g., language polishing) rather than to shaping the intellectual substance of the review.
Finally, we also recognize the accelerating presence of AI tools in qualitative research, particularly for tasks such as transcription, data management, coding assistance, and language refinement. These developments have prompted vigorous scholarly debate about the epistemological, ethical, and practical implications of integrating AI into interpretive work (e.g., Braun et al., 2024; Nguyen & Welch, 2025). While some view AI as a useful augmentation to human analysis, others caution that uncritical adoption risks undermining reflexivity, transparency, and interpretive depth. At GOM, we see this as an evolving space. We do not prescribe detailed rules at this time, but we expect authors to disclose how AI tools were used in the research process and to reflect on their implications for analytic validity and researcher agency. We encourage reviewers to evaluate such work with intellectual curiosity and methodological rigor, avoiding both uncritical acceptance and blanket rejection. As norms evolve, further guidance will likely emerge; for now, we urge authors to use AI thoughtfully, transparently, and in ways consistent with their stated epistemological commitments.
Conclusion
In closing, GOM’s identity rests on rigorous scholarship that takes groups and organizations seriously and on a developmental ethos toward authors and reviewers. For authors, we hope this editorial helps you decide whether a given manuscript fits the journal’s aims and scope, and how to sharpen your theoretical contribution, align it with your methods, and choose the right format. For reviewers, we hope it encourages you to calibrate your expectations to GOM’s mission rather than to a generic template, to avoid QRPs in reviewing—especially those that pressure authors into HARKing or chasing significance—and to evaluate each of our formats in a way that is fair, rigorous, and sensitive to its purpose. If we collectively hold ourselves to these standards—ambitious theory, appropriate methods, integrity in both research and reviewing, and respect for the diversity of our formats—GOM will continue to be a home for work that not only gets cited, but also changes the way we think about groups and organizations.
Footnotes
Acknowledgements
The initial version of this editorial was drafted by Yannick Griep and Billy Obenauer. All Associate Editors contributed to the development of the final version. Yannick Griep and Billy Obenauer are shared first authors; they are followed by Senior Associate Editors (alphabetical order) and then Associate Editors (alphabetical order).