David Ruppert and Nick Jewell are co-editors of the Journal of the American Statistical Association Theory and Methods section. We asked them to offer advice for prospective authors and discuss their thoughts about JASA and journal publishing in general.
Describe your roles as co-editors.
Perhaps our most important role was selecting the editorial board. We inherited an excellent board from the previous editors, but many associate editors felt they needed to step down for various reasons. We added a substantial number of young and mid-career associate editors who have done great work for JASA. We have weekly conversations generally on topics that range from individual papers to invited discussion topics and authors to assessing how JASA is doing in serving the statistical community.
Tell us a little bit about the process from submission to decision.
The entire review is conducted using the ScholarOne journal management system. Submitted papers first go to the editorial coordinator, Jamie Hutchens, to be checked for correct formatting and blinding. After Jamie has verified that a paper is ready for review, it goes into a “folder” of papers that need to be assigned to an editor. Each of us can either take a paper ourselves or assign to the other.
The paper then gets an editor screening review. About one-fourth of papers are rejected at this point. The most frequent reason for rejection is that a paper is narrowly focused and better suited to a more specialized journal.
Papers that pass the screening review are assigned to an associate editor. Selecting the best associate editor might be the most important job for an editor. Associate editors will sometimes reject a paper, themselves, often because the paper seems too incremental and sometimes because it is poorly written and difficult to read. Papers that are not rejected by the associate editor go to two, or sometimes more, referees.
Once the referees’ reports are in, the associate editor writes a report to the editor and sometimes to the authors making a recommendation. Frequently, the reviewers are in agreement and the editor’s job is easy. Occasionally, the editor needs to evaluate conflicting opinions very carefully before reaching a final decision.
Do you have any advice for prospective authors?
Authors should be realistic and not expect all, or even most, of their papers to appear in top journals. We urge authors to submit only their best work to JASA. Also, many papers are submitted prematurely, before the ideas are fully developed and clearly explained.
Before submitting a paper, authors should think carefully about the most appropriate journal. Perhaps our biggest surprise has been the poor quality of some papers whose authors believe they are suitable for a journal with the stature of JASA.
What are your top three pet peeves when it comes to submissions?
- 1) Researchers who agree to referee a paper, but never submit a report or even respond to emails from the editor or associate editor
- 2) Excessively long papers and unsolicited revisions
- 3) Authors exploiting the reviewing process to improve their work, rather than making sufficient effort up front
Certainly, the reviewing process is intended to improve papers and does that admirably, but a review should start with the best paper the authors can produce on their own.
During your time as JASA co-editors, can you point to any articles or issues you are especially proud of?
JASA-Theory & Methods receives 760 new submissions each year. It’s been difficult to get to know many of the articles well. We are pleased with the discussion papers that have been appearing regularly, including the invited paper presented at JSM each year.
What are the biggest challenges you face in your role as Theory and Methods editor?
Dealing with the very high submission rate has been a major challenge. Each of us was editing an average of one new submission a day.
What are your personal visions for JASA?
We think JASA should publish only papers that focus on broad statistical issues and provide innovative methodology motivated by a real application. We have striven to take JASA to new ground where possible and not simply mine old topics.
What do you think needs to change in journal publishing?
A fundamental challenge currently is the need to establish reproducibility of methods and analyses, an issue facing all scientific journals. Perhaps the field of statistics has special responsibility in this area, given that much of what we write about includes estimation algorithms and software code supplemented by relevant data analyses. We have pushed authors to supply code and data as supplementary files for all published papers wherever feasible, and now JASA is taking more formal steps in this direction under the guidance of Montse Fuentes. A bolder step would be to exploit current technology to have software built into the article, itself, but this requires digital, rather than publication.
Open access issues remain important, particularly as statistics and data science ideas are being shared globally.