International Journal of Eating Disorders: Statistical Guidelines

Cynthia M Bulik; Robert M Hamer; Ross Crosby; Tammy L Root; Ann Von Holle; Laura Thornton

doi:10.1002/eat.20677

. Author manuscript; available in PMC: 2013 Mar 15.

Published in final edited form as: Int J Eat Disord. 2009 Apr;42(3):193–194. doi: 10.1002/eat.20677

International Journal of Eating Disorders: Statistical Guidelines

Cynthia M Bulik ^1,², Robert M Hamer ^1,³, Ross Crosby ⁴, Tammy L Root ¹, Ann Von Holle ¹, Laura Thornton ¹

PMCID: PMC3598601 NIHMSID: NIHMS97130 PMID: 23505327

Introduction

In 2008, at the annual meeting of the International Journal of Eating Disorders editorial board, a discussion was begun that launched an effort to improve and standardize statistical reporting in the Journal. A statistical guideline committee was formed to explore the problem and develop a solution. Our volunteer committee comprised editorial board members as well as outside statisticians who are regular contributors to the eating disorders field.

After reviewing the state of the Journal, and canvassing both authors and reviewers, and surveying approaches to standardization that have been taken by other journals, we developed the 2008 IJED Statistical Guidelines. The resultant guidelines are designed to assist both authors and reviewers, to improve the standard of research presented in the Journal, and to provide a service to readers seeking consistency in reporting. Our first task was to develop formatting requirements for articles submitted to IJED. As requirements, authors are responsible for ensuring that their submitted manuscripts comply with the new IJED formatting rules. These requirements are consistent with current standards of statistical practice and will create consistency both within and across articles. Our second task was to develop statistical reporting guidelines. Although these are indeed guidelines and not requirements, authors should attend carefully to the guidelines and be aware that reviewers will be considering these principles when reviewing submissions to the Journal. For authors, attending to these guidelines in the design phase of any investigation will assist with ensuring a high quality scientific study. The topics that we chose to focus on include: one versus two tailed hypothesis testing, multiple comparisons, effect sizes, power analyses and sample size, pilot studies, misinterpretation of nonsignificant hypothesis tests, artificial categorization of variables, variable selection techniques, covariate adjustment, model validation, interactions, moderation and mediation, subgroup analyses, testing group differences at baseline, analysis of repeated measures, and missing data. In each instance, we supply a description of the issue and the rationale for the guideline, a succinctly stated guideline, and references to support the guideline.

The 2008 IJED statistical guidelines are detailed in a document linked to the Instruction to Authors page on the journal’s web site, http://www3.interscience.wiley.com/journal/34698/home/ForAuthors.html . We view this as a continuously evolving document. Statistical procedures are constantly evolving and so should our guidelines. We welcome comments on the 2008 guidelines and suggestions for further additions.