Editorial

The ‘Editorial’ article from the 2019-1 issue.

This article is converted from a Legacy LaTeX article using the texor package. The pdf version is the official version. To report a problem with the html, refer to CONTRIBUTE on the R Journal homepage. This article includes tables which may not be properly formatted. This article includes figures which have not been given correct alternative text.

The editorial board and I are pleased to present the latst issue of the R Journal.

We apologize that this issue has been so late in publication. As this is my first issue as Editor-in-Chief, I must personally thank Roger Bivand and John Verzani, the two previous EiCs, for their guidance in the technical aspects of putting an issue together.

The good news, though, is that publication should be much more timely in the future, due to improved internal technical documentation and the hiring of the journal’s first-ever editorial assistants, Stephanie Kobakian and Mitchell O’Hara-Wild. We are thankful to the R Consortium for a grant supporting the assistants (https://rjpilot.netlify.com).

This issue is chock full of interesting papers, many of them on intriguing, unusual topics. For those of us whose connection to R goes back to the old S days, it is quite gratifying to see the wide diversity of application areas in which R has been found productive.

Regular readers of this journal are aware of a change in policy that began January 2017, under which we are moving away from a paradigm in which a typical article is merely an extended user’s manual for the author’s R package.

To be sure, most articles will continue to be tied to specific packages. But we hope for broader coverage, and even the package-specific articles should emphasize aspects such as technical challenges the package needed to overcome, how it compares in features and performance to similar packages, and so on. As described in the announcement:

Short introductions to contributed R packages that are already available on CRAN or Bioconductor, and going beyond package vignettes in aiming to provide broader context and to attract a wider readership than package users. Authors need to make a strong case for such introductions, based for example on novelty in implementation and use of R, or the introduction of new data structures representing general architectures that invite re-use.

Clearly, there is some subjectivity in assessing these criteria, and views will vary from one handling editor to the next. But this is the current aim of the journal, so please keep it in mind in your submissions.

We wish the journal to further evolve in two more senses:

• In 2016, the American Statistical Assocation released a dramatic policy statement, seriously questioning the general usefulness and propriety of p-values. Though the statement did not call for a ban on the practice, it did have a strong theme that p-values should be used more carefully and less often. Many of us, of course, had been advocating a move away from p-values for years. We wish authors of future submissions to the journal to be mindful of the ASA policy statement. We hope for reduced emphasis on hypothesis testing, and in articles that do include testing, proper consideration of power calculation.
• In the interest of reproducibility—a requirement already imposed by the journal on article submissions—we will require that any real datasets used as examples in an article must be provided. Note that this will mean that datasets with privacy issues or datasets of extremely large size should not be used in an article.

Finally, we note our deep appreciation for the anonymous reviewers. A journal is only as good as its reviewers, and most reviews are quite thoughtful and useful. If a handling editor solicits your review for a paper, please make some time for it. And if you must decline the request, a reply to that effect would be quite helpful; don’t just discard the editor’s e-mail message. The handling editors are quite busy, and it is unfair to both them and the authors to have the editors wait until they must conclude you will not reply, causing unnecessary delay.

Note

This article is converted from a Legacy LaTeX article using the texor package. The pdf version is the official version. To report a problem with the html, refer to CONTRIBUTE on the R Journal homepage.