The Collaborating Statistician: Publishing in a Peer-Reviewed Journal

Alexandra HanlonAlexandra L. Hanlon has been a research professor in the University of Pennsylvania School of Nursing since 2009. As the founder and faculty director of the BECCA Lab, Hanlon has enhanced the school of nursing’s scholarly mission by expanding its capacity and expertise in statistics.

As a University of Pennsylvania School of Nursing faculty member, I facilitate faculty and student research endeavors as a collaborative biostatistician. Through this role, I have published widely and gained considerable external funding. I am an active participant in various professional committees at the local, regional, and national levels. Serving on doctoral committees and directing independent studies, research residencies, and research assistantships, I provide a range of mentoring support. Also, I formally direct and mentor graduate interns of biostatistics and epidemiology through our Biostatistics, Evaluation, Collaboration, Consultation, and Analysis (BECCA) Lab.

BECCA serves students throughout the Philadelphia region in such programs as public health, applied statistics, biostatistics, and epidemiology; they are seeking experience as health science collaborators. They become members of multidisciplinary teams and contribute as collaborative applied statisticians under the mentorship of senior statisticians. On these teams, they improve their written and verbal communication skills—along with negotiation skills, time management, professional responsibilities, and appropriate initiatives. Many collaborate on published papers and thereby begin to build their professional résumés.

This article provides a flexible, step-by-step guide for statisticians who are collaborating in crafting a manuscript for a peer-reviewed clinical journal.

First, the team should agree on the end goal of the inquiry. This goal should address a solid research question and, if appropriate, be based on a theoretical framework. Next, the team members should agree on each member’s role in the project. This agreement should include the team’s order of authorship. This tricky business could be facilitated by guidance provided in the International Committee of Medical Journal Editors (ICMJE) Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. The team should then select the targeted journal for publication.

Selecting a Journal

The targeted journal should be a good fit for the study’s depth and focus. Does the study have broad scientific interest, or a narrower focus? Are the study’s findings incremental, or do they portend a breakthrough of the current paradigm? Faulty targeting can seriously delay the publication process. Journal editors are averse to multiple submissions, so manuscripts must be submitted sequentially.

After a journal has been targeted, the lead author should submit a brief letter or email to the journal’s editor to determine the level of interest in the manuscript’s title and summary. Following a favorable response, the team should begin composing the manuscript. In the case of an unfavorable response, the team should reconvene, consult with experienced colleagues, and then select a second journal. The team may also consult JANE: Journal/Author Name Estimator, a free web-based tool that helps identify suitable journals. JANE also suggests relevant articles to cite in the team’s manuscript. After selecting a journal, the team should carefully review and follow the journal’s guidance on format, style, length, language use, and document design for tables, figures, and appendices.

Composing a Manuscript

The team is now prepared to compose its manuscript for submission to the selected journal. The lead clinical investigator should prepare the introduction, the methods and materials section, and the discussion. However, space should be left for the statistical methods section, which will be drafted later by the collaborating statistician. The introduction should begin with a background description of the current state of the scientific field, supported by references to works central to the field. It should offer reviews along with key original scientific reports. It should then identify the gap its study will fill in this scientific field. It should clearly establish the study’s focus and describe the study’s significance. The introduction should conclude with a statement of the goal of the study. The following section on methods and materials should be comprehensive and offer sufficient details to enable replication of the study.

The collaborating statistician should be responsible for laying out the tables and figures. They should be presented in a way that “tells the story” logically and succinctly while directly correlating with the goal of the study. This layout should be used to organize the statistical analysis methodology and results sections. Each table or figure should make a single clear point as reflected by its title. Legends should be written such that the tables or figures stand alone and are understandable in isolation. The final manuscript should present these in high resolution. Blurry, hard-to-read figures detract from the text and go unheeded by many readers. The figure panels should be arranged to flow with the story of the study.

The collaborating statistician should then draft the study’s results in a sequence of subsections that match the titles or legends of the tables and figures. Each subsection should conclude with a summary sentence that enables readers to synthesize the study’s findings. The lead clinical author should then edit the collaborating statistician’s results section to ensure that technical language is softened and more accessible to a broader clinical audience. The collaborating statistician should now compose a description of the statistical methods that produced the results of the study. The comprehensive description should provide sufficient detail to enable replication of the study.

After the study’s results have been summarized, the lead author can describe the study’s place in the larger scientific context in the discussion section. This scholarly discourse affords the team a place in the greater research community; they too can be cited. Their work may advance the field, or it may open the way to research areas that were heretofore unexamined. At this point, all the study’s references should be noted and scrupulously documented. Some journals require a limitations section that specifies sub-optimal aspects of the study. The lead author should note such limitations, with the team’s inputs. If the targeted journal allows appendices, they provide a convenient place to attach related complex supporting tables or figures to your submitted manuscript.

Using the results section as a guide, the lead author can now compose the article’s abstract. Usually strictly limited to a couple hundred words, the abstract distills the study’s findings. The team should then agree on an accurate, concise title. The article will be indexed by its title, which should attract citations. It should capture the main points of the paper and reflect the style of other titles in the journal.

The complete draft is now ready for circulation among the team, other contributors, and trusted first readers. They should be encouraged to offer feedback within a couple of weeks. After a final edit in response to this feedback, the lead editor should circulate the draft again to all interested parties for their final review.

Submitting a Manuscript

After all coauthors have scrutinized the final draft, they should notify the lead author of their consent to publish. The lead author should make a final check that the submitted draft complies with the journal’s format, policies, and procedures. The lead author should then upload the finished manuscript with a cover letter and all supplemental materials. The cover letter should briefly highlight the main points of the manuscript, cite its importance, and explain its appropriateness for publication in the selected journal.

A positive response following peer review typically requests changes for the published article. The team should quickly consider theses changes, noting that these suggested final revisions almost always improve the final product. The team should then forward their final text to the journal for publication. If accepted, most journals will provide their authors with a galley proof of the text that will be published. The team should expeditiously do one last proofread of their article, taking advantage of a final opportunity to make very minor changes to the text.

If the selected journal rejects the initial submission, the team should review the editor’s reasons for rejecting the manuscript. The rejection letter may include peer reviewers’ reservations. The team should then decide whether they should revise and resubmit their manuscript to another journal. Of course, a resubmission typically requires adapting the manuscript to the new journals’ format and style guide.

Unpublished authors should find comfort in knowing that rejection refines us. It is not personal, and those who are persistent will survive the arduous peer-review process.

Print Friendly