Pre-analysis planning

In general, we warmly encourage our authors whose research designs allow it to engage in transparent and traceable pre-analysis planning. For research relying on hypothesis testing and when novel data are being collected, we encourage authors whenever possible to pre-register their hypotheses/expectations. Especially for quantitative articles (but not exclusively), we encourage authors to develop pre-analysis plans (PAPs) and register them along with their hypotheses. Concretely, authors submitting to EJPR are encouraged to (i) pre-register their expectations and pre-analysis plans, unless it is impossible for them to do so, and share these materials anonymously upon initial submission. Several independent, free organizations allow researchers to pre-register expectations and procedures, for example, the OSF framework; we encourage the authors to make use of such services, (ii) adhere to pre-registered procedures, conduct all tests that were part of pre-analysis plans, and test pre-registered hypotheses, (iii) mark analyses not included in the pre-analysis plan as exploratory, and (iv) discuss explicitly any type of deviation from pre-registered procedures and expectations. Such deviations may be well-motivated and do not necessarily signal issues with procedures or flaws of the pre-analysis plans themselves (eg Lakens Reference Lakens2024). We do not see pre-registrations and pre-analysis plans as setting immovable guardrails for what can and cannot be done, but simply as instruments to encourage researchers to be upfront and transparent about intentions and subsequent actions. In case of any deviation, authors are expected to include an explicit discussion about these deviations, their reasons, and possible impacts (for instance in a deviation table in the supplementary materials). Our editorial team makes a firm commitment not to publish articles based on how ‘successful’ the results of the analyses are and fully considers null-result articles as an equally valid form of scientific output as articles where hypotheses are supported by empirical evidence. Similarly, we commit to fully consider analyses with explicit and transparent deviations.

Registered reports

Expanding and formalizing the logic and philosophical underpinnings of pre-analysis planning, in 2027 the EJPR will introduce a new type of article type – registered reports (RRs; Chambers and Tzavella Reference Chambers and Tzavella2022). In a nutshell, RRs ask authors to submit their research design (hypotheses, analysis plan) before they proceed with the empirical analysis per se, which is what undergoes peer review. By evaluating the research design before results are known, RRs are expected to reduce well-known publication biases such as the preference often given to significant or positive results (thus effectively contributing to reducing the ‘file drawer’ problem; Laitin, Miguel, Alrababa’h et al. Reference Laitin, Miguel, Alrababa’h, Bogdanoski, Grant, Hoeberling and Williamson2021) or tailoring theoretical expectations and scientific contributions post hoc, based on the nature of the results found (HARKing; Kerr Reference Kerr1998; Lakens, Mesquida, Rast et al. Reference Lakens, Mesquida, Rasti and Ditroilo2024). Additionally, encouraging scholars to think ahead about their analysis plans reduces researchers’ degrees of freedom and p-hacking (Arpinon and Espinosa Reference Arpinon and Espinosa2023), which likely makes published research more transparent and replicable. Preliminary evidence from research in psychological and neuroscience shows that, compared to similar ‘traditional’ articles, registered reports contain research that is not less novel or creative – contrary to a critique that is sometimes ascribed to pre-planned research – but include research that is much more rigorous from a methodological standpoint. And while registered reports are still vulnerable to manuscript editing done simply to ‘please’ reviewers, for instance altering the nature of the original theoretical framework based on the suggestions received even if they disagree with them (‘dishonest conformity’; Shibayama and Baba Reference Shibayama and Baba2015), they nonetheless ensure that this is not done on the basis of the results found (eg the so-called TARKing, or ‘theorizing after results are known’; Weiss, Doden, Antino et al. Reference Weiss, Doden, Antino, Schmutz and Unger2026).

Against the backdrop of a decreasing trust in science and other epistemic authorities (van der Meer and Hameleers Reference van der Meer and Hameleers2026), and in light of several high-profile scandals about scientific fraud and data tampering, registered reports can operate as a bulwark to boost trust in empirical results produced by the scientific community (but see, for a more cautious take, Costa, Inbar and Tannenbaum Reference Costa, Inbar and Tannenbaum2022).

Concretely, in RRs authors first submit the equivalent of an introduction with framing and contribution, a theoretical rationale, and a detailed pre-analysis plan, before any data are collected (aka, ‘stage 1’), which undergoes peer review. Reviewers evaluate the soundness of the theoretical and empirical frameworks presented and decide whether the submission meets the standards of impactful research. Based on the reviewers’ recommendations, the editor decides whether to grant the authors an ‘in-principle acceptance’ of their article (either immediately or after a round of revisions, like during ‘normal’ peer review). If so, the authors are invited to collect the data and proceed with the empirical analyses in line with what they have described. Their results are then presented in the finalized article (aka, ‘stage 2’), which is resubmitted and checked by the reviewers against what was promised during stage 1. The article is then published as is.

It is important to note that registered reports are not mandatory. They are offered as an alternative submission format and should not be seen as forcing empirical research into a rigid or constraining format. Authors are, and will always be, free to submit traditional empirical articles. Additionally, RRs can be tailored to all types of empirical frameworks that follow a clear hypothetico-deductive logic. This includes empirical frameworks relying on large-N quantitative data from surveys, (survey) experiments, and other large matrix datasets based on figures or text. It also includes empirical frameworks relying on small-N qualitative data from interviews or documents, when these data are analysed using a deductive logic. However, RRs are not suited for inductive approaches using ethnographic or interpretative methods.

A new reproducibility policy

According to recent research, only one quantitative article out of three (31%) published in major political science journals makes the data available for replication (Rainey, Roe, Wang et al. Reference Rainey, Roe, Wang and Zhou2025). We believe we could do better, and that as editors of a major journal in the discipline, we have a responsibility to do our part. The EJPR is committed to the highest standards of research transparency and integrity. With this in mind, we have designed a new reproducibility policy to achieve three core objectives: (i) Transparency. The procedures used to collect and analyse data, as well as the data itself, should be as fully available as possible to enable future researchers to build upon, critique, and extend the published work. This applies to both quantitative and qualitative research. (ii) Comprehensibility. Readers with the appropriate technical knowledge can understand how a paper’s findings were produced by reviewing the manuscript, appendices, and the accompanying reproducibility package. (iii) Quality Control. Empirical results published in the journal should be computationally sound and free from error. We are currently envisioning replicating a random selection of articles every year, systematically reproducing the analyses of the selected articles.

Concretely, to comply with these guidelines, authors for all articles that rely on statistical or computational methods – including but not limited to quantitative analysis as well as qualitative research that uses computational techniques (eg automated content analysis, Qualitative Comparative Analysis) where data are organized in a matrix form – will be asked to submit a ‘reproducibility package’ to be approved by the journal before articles proceed to copy editing.

The guidelines contain both required (eg share data and code allowing to replicate the main results in the article) and recommended (eg sharing the original, unprocessed dataset) courses of action. Authors submitting to EJPR commit to these guidelines, and compliance is a prerequisite for publication. Exceptional exemptions can be granted upon submission, if authors explain why these guidelines are not applicable for them.

Acknowledging roles and responsibilities

Collaborative research is increasingly common in the social sciences, including political science. Data from submissions to the EJPR show that the average number of authors per submitted article has been steadily increasing over the past decade, from 1.8 authors in 2017 to 2.2 in 2025, following the general trend in the discipline (Fisher, Cobane, Vander Ven et al. Reference Fisher, Cobane, Vander Ven and Cullen1998; Grossman, Dinneen and Torreblanca Reference Grossman, Dinneen and Torreblanca2025). Given the continued presence of single-author submissions, this indicates that very often articles are authored by 3 or more scholars. In most cases, collaborative research is conducted in a cooperative and collegial manner, but conflicts between authors – including disputes about work done, misconduct, and even abuse of power when early-career scholars are involved – is unfortunately not something that can be excluded (eg Cohen and Baruch Reference Cohen and Baruch2022). To ensure that each author receives the credit that is due to them and generally to foster accountability of all research involved, it is important to know ‘who did what’.

Articles published in the EJPR will be henceforth warmly invited to include a contribution statement specifying the role of each co-author (eg funding acquisition, project management, writing, editing, and data collection), building on the framework developed by the Contributor Roles Taxonomy (CRediT), effectively ‘shifting from authorship to contributorship’ of our published articles (Allen, O’Connell and Kiermer Reference Allen, O’Connell and Kiermer2019).

General editorial policy

EJPR is committed to scientific excellence, and our editorial decisions should reflect this overarching commitment. Therefore, we uphold the following norms for editorial decisions: (i) the Editors assess the general sentiment of the reviews collected. They respect and value the insights provided by the reviewers and decide accordingly; (ii) the minimum requirement for granting the authors the possibility to revise their manuscript (R&R) is usually that at the very minimum one set of comments expresses a positive sentiment. The bar to reach the R&R stage is thus high and requires some form of preliminary support from all reviewers for the article upon first submission. If all reviewers are (very) critical and suggest considerable or hard revisions, the article will likely be rejected for insufficient initial support; (iii) it is possible for Editors to override the general sentiment expressed by reviewers (eg the general sentiment is negative, but a R&R decision is nonetheless made), but only in the presence of strong reasons; (iv) ideally, we strive to collect at least three sets of reviews, to ensure more diverse and complementary insights. Reviewers are usually a mix of topical and methodological experts, depending on the nature of the manuscript being evaluated; (v) Editors reserve the right to make editorial decisions based on a smaller subset of reviewers, in case the comments collected are very clear and converging, and the review process has been excessively long; (vi) in case of doubts about the general sentiment, Editors ask for a second opinion by a fellow Editor, and decide accordingly; (vii) if one Editor suspects the presence of fraud (eg plagiarism), they alert the Editorial team about possible plagiarism/fraud. At least one other Editor must agree that there is plagiarism/fraud. If so, the article is (desk) rejected.

Multiple revision rounds

An increasingly common practice in social science journals is to put papers through multiple rounds of peer review and subsequent substantial revisions, often more than two. While in some cases this is warranted, we strive, as much as possible, to make an editorial decision after one round of revisions. Very concretely, this implies the following: (i) if an R&R is granted, authors get one chance for a major revision of their article. A final decision will be taken after this round of revisions, in consultation with the original reviewer(s) (and possibly additional ones); (ii) if one or more reviewers remain (or are) sceptical after assessing the revised article, the article will likely be rejected for publication. In practice, this means that a second decision of ‘major revisions’ will only very exceptionally be granted. After a first round of revisions, articles will either be (conditionally) accepted or rejected. Reducing the number of rounds has several positive consequences: it shortens the peer-reviewing process, clarifies the process for the authors, and reduces the overall burden for reviewers.

Taking desk rejects seriously

Our commitment to publish scientific excellence and the stricter editorial norms discussed above imply that desk reject is a very common occurrence when submitting to the EJPR – the statistical mode. Beyond upholding high standards of what goes under review, desk rejects contribute also to keeping the burden put on reviewers – who always donate their time and energy voluntarily and most often receive little recognition for their efforts – as low as possible.

We commit to provide authors of desk-rejected manuscripts with a substantive justification about the reasons why their article was not sent out for review, and suggestions to improve it. We strive to avoid using template ‘desk reject’ emails, and instead ensure that authors benefit from their submission, even if the outcome is negative. Desk reject is not a pleasant outcome for authors, but it can still be a formative experience.