Publishing in the BMS. A long process,but not without reason

The latest issue – 161-162 – has just been published at the time of writing. Of the eight articles published, one was submitted less than six months ago, in autumn 2023; five were received about a year ago, between March and May 2023; and finally, two articles have been awaiting publication for much longer: one was submitted in July 2022, the other in March of the same year. Unfortunately, these two articles were written by the youngest, non-tenured authors, for whom this delay is obviously detrimental. On the other hand, the article that was published quickly was written by the most advanced author in his career. We do not have the means here to objectively measure the time it takes our colleagues to publish in BMS, but we do know that it seems a long time to them, just as it seems a long time to us to publish in other peer-reviewed journals.

What are the reasons for these delays? Are they so unreasonable? To understand them, let’s look at the various stages in the publication process of articles in our journal – and in most journals that use external double-blind review.

The first stage is called ‘deskreject’: when an article is submitted, the ScholarOne platform informs the editors – in this case the four people who signed this editorial. One of us, volunteering our skills and availability, reads it in detail. Unlike other journals, we don’t just read the abstract but the whole text. This means devoting several hours to this initial approach to the text. Two criteria are used to decide whether or not to send the article for review. First, is the text in line with the editorial policy of the journal? The BMS has defined an original scientific policy in the field of international methods journals: ¹ we avoid ‘discourses on methods’ and instead promote reflective accounts of experiences. But is the text of sufficient quality to justify the time we are asking potential reviewers to devote to it? The editors discuss the proposal to reject or send for review before the decision is communicated to the authors. In some cases, we ask for a new submission that better meets our expectations before sending it for review. The ScholarOne platform sets a deadline of one week for the deskreject, but because of our respective commitments, it usually takes us between two and four weeks to complete this stage, especially as articles often arrive in waves, particularly on vacation eves.

For those articles that have not been rejected outright, we then have to choose the reviewers to whom they will be entrusted. We won’t go into the principles that guide our choice here, nor into our expectations of these reviewers. We’ll devote our next editorial, in issue 164, to this subject. What we do want to highlight here, however, is the difficulty of getting two colleagues to agree to carry out this task, hence the delay. Again, BMS is far from being the only journal to encounter this difficulty, which seems to have increased considerably since the pandemic. Double-blind review is an invisible and, in social sciences at least, unpaid labor. It is very useful to the scientific community in general, and to authors and publishers in particular, but very time consuming. Many reviewers refuse our invitation. Getting two positive answers from the reviewers at the first attempt is exceptional. Again, many thanks to those who agree to our requests. Colleagues who refuse are more likely to do so by not responding. The process of identifying potential reviewers, finding their contact details, contacting them, getting their agreement and possibly negotiating the deadline can take several additional weeks. Not to mention that it’s not uncommon for a colleague who has agreed not to follow up without necessarily telling us. ² Once the statutory two months have passed and several reminders have been sent out, the process starts all over again. If all goes well and the reviews were not too difficult to obtain, it takes at least two months to compile the reviews, but this time is easily doubled by the above-mentioned difficulties.

It should be noted that the procedure described so far applies only in part to articles proposed by colleagues as part of a special issue or a guest section. ³ Since these articles have already been the subject of preparatory work by the guest editors, the deskreject process is a residual one. In this case, we only ask for one external reviewer instead of two. Another advantage is that special issues proposals are also filtered by the journal’s editorial board, so the dossier editors benefit from an additional body of comments and suggestions. The articles in the ‘The making of my research’ section, which we solicit from well-known scientists, are also largely exempt from this procedure: they are not double-blind, but are commented on directly by the journal’s editorial board. ⁴

When the reviews are available, we use them as the basis for the decision to be sent to the author(s). These decisions are always substantiated. We don’t just keep score and pass on reviews as they come in. The person in charge of the article reads the reviews and prepares a response. All situations are possible. Sometimes the reviewers agree with each other, but by no means always. The content of the reviews is sometimes substantial, sometimes minimal. In most cases, we are far from being able to simply accept or reject an article on the basis of the reviews alone. We are therefore often called upon to arbitrate and supplement the opinions received with our own assessments or with those of members of the editorial board. Again, it takes two weeks to a month, or even a month and a half if the reviews are contradictory, to inform the author(s) of our decision. All in all, the whole process, from submission to the decision to publish – with minor or major revisions – or to reject an article after peer review, takes between three and nine months.

The fate of those articles that require only minor changes in order to be published is then sealed. Some of these changes are still of a scientific nature, relating to remarks and comments that were only partially taken into account by the author(s), but most of them are formal requests, relating to the de-anonymisation of the manuscript, the meticulous checking and correction of the bibliography, and compliance with the SAGE conventions on the presentation of the text and its authors.

On the other hand, very few articles are definitively rejected once we get feedback from the reviewers. Actually, it is very rare for both reviewers to recommend rejection, which is not surprising given that we carefully select the scientific level of the articles we submit to our reviewers. If at least one of the reviewers suggests that the text should be revised, we do so in most cases.

We always provide detailed reasons for our requests for changes, based on the reviews we have received and the expertise of the editorial team. If the two reviews go in the same direction, we highlight their convergence and stress the need to change the text accordingly. If the reviews differ from one another, we indicate which is our opinion and which is the opinion of the reviewers. It is not uncommon for us to make additional suggestions based on the coherence of our editorial project and our way of thinking about methodological research.

The article will then be revised. We leave it up to the authors to indicate when they think they will be able to complete the revision and send us the new version of their article. Again, deadlines vary widely. We don’t force things, as we are well aware that authors are no more available than we are ourselves. Each article then takes a different course. Some are thoroughly rewritten in a short space of time, in line with our expectations, and can be published quickly. Others have to go back and forth several times, in a process that can sometimes be painful. This is particularly the case for articles submitted by young colleagues ⁵ in the ‘The method of my thesis’ section. We are committed to not rejecting them at any stage of the reviewing/editing process, but this sometimes comes at the cost of three or four re-submission requests, which we know are difficult both materially and psychologically for young colleagues in the race for permanent positions. But the end result is much better.

The editorial team manages an average of a dozen articles at various stages of the process, not counting special issues in preparation. The attention we give to each article also depends on the journal’s schedule. We submit one issue (an average of five articles) per quarter. As soon as one issue is submitted, we start working on the next one, identifying articles that we think should be ready for publication in two months’ time. For these articles, their reviewers and their authors, history is accelerating. Just when they seem to have fallen off the radar, buried in the depths of ScholarOne, we pick them up again, and re-launch the reviewers and author(s). This change of pace may come as a surprise, but it accelerates until the text is submitted. Previously, we mainly asked authors to be patient, but at this stage we rely on their responsiveness and urge them to reply to our requests by email. We would like to take this opportunity to thank our colleagues for the friendliness and efficiency with which they generally respond to these requests.

From the moment the issue is sent to Emilie Bickerton, our project manager at SAGE, the deadlines become predictable. Three months would pass between the submission of the texts and the publication of the issue. Emilie checks our files, gets back to us when necessary and, most importantly, configures them for production. She liaises with the typesetters and sends us the first set of proofs about six weeks after the issue is submitted. We then have one week to send them to the authors and proof them ourselves. Emilie checks that the typesetters understand the changes we have requested and, if necessary, gets us a second set of proofs. About three weeks later, she sends us the proofs of the entire issue. We, the editorial team, then do a final reading and the issue goes to press. Two weeks later, it is usually an author who writes to us saying how happy they are to see their text finally published – or Karl van Meter who announces it on the RC33 mailing list.

Given this editorial process, are publication times of six months to two years unreasonable? They are certainly very costly for young authors who need publications to land the Grail and get a permanent position. Only very large journals with high impact factors can work with much shorter deadlines. This is partly because they rely on the work of publishing professionals, not scientists, to manage the review process; and partly because their visibility in their respective fields implies receiving a very large number of submissions, from which they can select only those articles that are practically ready for publication. For niche journals like ours, the added value of editorial work is as essential as it is time-consuming. By supporting authors in this way, we try to influence the way methods are used in our small field of social sciences, and in doing so, we make our own scientific contribution.

This issue contains six articles. It begins with a long article that Antoine Machut submitted in the section ‘the method of my thesis’, which explains how he studied the careers of financial journalists in France. He describes the empirical tools he used and how he combined them to identify the structure of these careers in a particularly uncertain professional world. The data collected is extremely rich and the methods used demonstrate, among other things, the value of using qualitative interviews based on biographical narratives to create a dialogue and to shed light on the results obtained by sequence analysis (optimal matching). Three narrative patterns emerge, combining the regularities of the reconstructed careers of 297 journalists and the interpretation of 37 biographical accounts. Discover this text with the last one in this issue, written by Sébastien Michon, who studies the careers of parliamentary assistants in Brussels and compares sequence analysis with other methods of creating typologies.

Two much shorter articles come next in the ‘Tools and instruments’ section. In the first one, David Longtin, Marie J. Bouchard, Valérie Michaud and Damien Rousselière review the methods used to analyse social innovation. In full expansion since the 1990s, work on this topic has often struggled to move away from descriptive case studies. To remedy this situation, the Centre de recherche sur les innovations sociales (CRISES) at the Université du Québec à Montréal created a database on social innovation that allows the statistical analysis of around one hundred case studies from Québec. The article discusses the originality, limitations and analytical scope of this methodology in relation to existing mappings of social innovation. In the second one, Piotr Jabkowski, Piotr Cichoki & Marta Kołczyńska focus on the impact on the quality of the response rate of two methods of selecting respondents to be interviewed within households in a questionnaire survey with a probabilistically constructed sample. In this case, they consider two methods that dominate international surveys of the general population: the Kish grid and the birthday method. They use the European Social Surveys (ESS), a recurrent international questionnaire survey conducted face-to-face with a randomly selected sample.

Three ‘Implementation’ articles follow. In the first one, Cédric Calvignac, Franck Cochoy and Gérald Gaglio present a method rarely used in sociology, the ‘call for testimonies’, which consists of collecting written accounts of the experiences of a large number of actors on a given social fact. The authors themselves tested the method by asking their fellow citizens to testify about the use of masks during the COVID pandemic. After using this survey extensively and even writing a manual in French on this original method, they explain what they have learned from the method and illustrate it with the self-production of masks at the beginning of the health crisis. Next, Jeanne Hersant, Andrea Vera Gajardo and Mariel Saez Trumper tell the story of a collaboration between sociologists and women mathematicians in Chile, aimed at understanding the specific nature of the careers of women mathematicians in that country, careers that are rare compared to men’s careers, but particularly successful. The authors discuss the issues raised by interdisciplinarity and feminist epistemology, before highlighting the hiatus that their collaboration enabled them to identify, particularly in terms of language.

Finally, echoing the first article in this issue on the careers of financial journalists, Sébastien Michon examines the careers of parliamentary assistants in the European Parliament. He develops a fascinating reflection on the value of carrying out several quantitative analyses based on different statistical schools and comparing them (in this case, the typologies obtained by means of a sequence analysis or a geometric analysis). The aim is to gain a better understanding of the blind spots of each of these methods and, through their similarities and differences, to question our sometimes too routine practices. Sébastien Michon invites us into his analytical kitchen, showing us the stages of his thinking and how he manages to take into account the time taken by trajectories in geometric analysis in order to better compare and enrich the interpretation of the typologies obtained.

Enjoy your reading!

April 2024