JAMA Network Guidance on Venn Diagrams

Connie Manno, ELS, Director, Freelance Editing Unit, JAMA Network

Venn diagrams are simple pictorial representations of relationships that exist between 2 or more sets of things. Circles that overlap have commonality; circles that do not overlap do not share traits.1

Although Venn diagrams represent conceptual shared or unique traits between separate ideas or groups of things (Figure 1), they are not appropriate to visualize numerical (empirical) data.2

Figure 1. Conceptual Venn Diagram

Often, the separate sets are presented as identically sized circles—even if the quantities in each set and the overlapping and nonoverlapping segments are different—and the resulting illustration can be not only imprecise but also misleading (Figure 2).3

Figure 2. Venn Diagram of Identically Sized Circles That Represent Different Quantities

The identically sized circles obscure the different numbers of cohorts included in the referenced studies. From Sentenac et al.3

Like pie charts, which also compare relationships among component parts and are frequently used to depict data for a lay audience, Venn diagrams should be avoided in scientific publications.4(pp137-138)

One more precise way to present the data is to create a bar graph or component bar graph (Figure 3), which can present the relationships between 2 or more data sets while illustrating the size difference between the sets with bars of unequal lengths. A component bar graph additionally uses color and section length to highlight patterns in the data.2

Figure 3. Data as a Component Bar Graph

When the data from Figure 2 are presented as a component bar graph, the difference in cohort sizes is apparent from the bar lengths. In addition, bar sections that depict shared segments use the same color.

Another option is to present the data in a matrix: a tabular structure that uses numbers, short words (eg, no, yes), symbols (eg, bullets, check marks), or shading to depict relationships among items in columns and rows and to allow comparisons among entries.4(p114)

Depending on the complexity of the construction and the need for multiple colors or shading, a matrix may be presented as a table or figure (Figure 4).

A third option is to resize the circles to make them more proportional to the quantities they represent (Figure 5), but only if the circles and overlaps are precise and generated from statistical software.

Figure 5. Circles From Figure 2 Resized to More Accurately Represent the Sample Sizes

This option must use precisely sized circles and overlaps generated from statistical software to ensure that the figure’s elements are truly proportional.

Network figures that use nodes and connecting lines of varying sizes to illustrate the proportions of the compared items are also useful for depicting relationships among 2 or more sets of data (Figure 6).5

Figure 6. Network Figure Depicting Relationships Among 4 Data Sets

In this network map, the size of the nodes is proportional to the number of participants in each node, and the thickness of the connecting lines is proportional to the number of randomized clinical trials in each comparison. From Ferreyro et al.5

Although data can be displayed multiple ways, accuracy and audience, as well as the criteria of the final format (eg, scientific journal vs consumer publication), should govern the decision of which option to use.

References

  1. Kenton W. Venn diagram. Investopedia website. Updated January 17, 2020. Accessed February 1, 2021. https://www.investopedia.com/terms/v/venn-diagram.asp
  2. Harris RM. Bar plots as Venn diagram alternatives. Rayna M. Harris blog. May 7, 2019. Accessed February 1, 2021. https://www.raynamharris.com/blog/vennbar/
  3. Sentenac M, Boutron I, Draper ES, et al. Defining very preterm populations for systematic reviews with meta-analyses. JAMA Pediatr. 2020;174(10):997-999. doi:10.1001/jamapediatrics.2020.0956
  4. Tables, figures, and multimedia. In: Christiansen S, Iverson C, Flanagin A, et al. AMA Manual of Style: A Guide for Authors and Editors. 11th ed. Oxford University Press; 2020:113-169.
  5. Ferreyro BL, Angriman F, Munshi L, et al. Association of noninvasive oxygenation strategies with all-cause mortality in adults with acute hypoxemic respiratory failure: a systematic review and meta-analysis. JAMA. 2020;324(1):57-67. doi:10.1001/jama.2020.9524

Welcome the 11th Edition of the AMA Manual of Style!

We are pleased to announce the 11th edition of the AMA Manual of Style, now live at https://www.amamanualofstyle.com/ and shipping in hardcover in a few days.

The manual has been thoroughly updated, including comprehensive guidance on reference citations (including how to cite journal articles, books, reports, websites, databases, social media, and more), an expanded chapter on data display (for the first time in full color), a completely up-to-date chapter on ethical and legal issues (covering everything from authorship and open access to corrections and intellectual property), and updated guidance on usage (from patient-first language and terms to avoid to preferred spelling and standards for sociodemographic descriptors).

The section on nomenclature has undergone thorough review and updating, covering many topics from genetics and organisms to drugs and radiology.

The statistics and study design chapter has been extensively expanded, with more examples of usage and terms that link to a related glossary.

Chapters on grammar, punctuation, abbreviations, capitalization, manuscript preparation, and editing feature refreshed examples and new entries (such as allowance of the “singular they”).

The nearly 1200-page book is enriched by a variety of online features. For example, regular updates to address changes in style or policies will be featured in the Updates section. Any corrections will be made online so that you are always looking at the latest guidelines as you use the manual.

New quizzes will be posted to help new or continuing users learn to master the finer points of AMA style, and the units of measure calculator offers easy conversions between the SI system and conventional units, as well as the metric system.

We welcome questions and comments on the manual: write to stylemanual@jamanetwork.org or find us on Twitter (@AMAManual). We look forward to engaging with you. –Stacy Christiansen, for the AMA Manual of Style Committee

Looking Forward to the 11th Edition

At the most recent annual conference of the American Medical Writers Association, we received a sneak peek at some of the changes to come with the AMA Manual of Style 11th edition. AMA Manual of Style Committee Members Cheryl Iverson, Stacy Christiansen, and Annette Flanagin gave an overview of some highlights, including changes to the way references are styled, updates taking into account the growing presence of social media, and changes to nomenclature. Other topics included updates to corrections processes, clarification of authorship guidelines, and data sharing rules.

In this post, I thought I’d share some of the changes that I’m most excited about as an editor, many of which promise to make the editing process easier and more intuitive. Several new guidelines for styling references caught my attention. For example, electronic references in the current edition mirror traditional references with the addition of a URL in the middle of the reference and a date accessed at the end. The new edition will put URLs at the end of a reference, similar to how DOIs currently appear. Another welcome change is the omission of publisher locations from references to books and reports, which seems sensible given the multinational nature of publishing and the increasing move toward digital formats.

An update to the style of tables and figures was also covered in the presentation. Starting with the 11th edition, column heads in tables and axis labels in graphs will appear in sentence case capitalization rather than title case. Table and figure headings and labels in scientific publishing often contain a great deal of text, making it all the more important to display that text in a way that’s as easy as possible for readers to parse. In addition to saving space, the use of sentence case capitalization throughout figures and tables will make these elements appear more standardized and easier to read.

The new edition also makes some helpful distinctions between AMA style preference and alternatives considered acceptable. For example, it is now acceptable to use of “they” as a singular pronoun. In addition to circumstances in which “they” is a person’s preferred pronoun, this construction can also be helpful in preserving patient confidentiality (eg, avoiding the disclosure of patients’ sex in a small sample size to preserve anonymity).

Many other exciting changes are anticipated with the upcoming 11th edition. Stay tuned for more!—Heather Green

 

 

Get to the Point!

Here comes Hank. Too late, he’s spotted you, and now you’re in for another story—or rather, a litany of unnecessary details. “I said this, and she said that, and then I said, ‘Really!’” Hank never edits himself; he simply tells you E-V-E-R-Y-T-H-I-N-G until you’re screaming inwardly, “Get to the point!”

While editing manuscripts, I periodically encounter a “Hank” author. Every tidbit of information is important and, in his view, absolutely necessary. Along with his manuscript, which includes the maximum-allowed 5 tables and/or figures, he provides a Supplement that comprises 3 eMethods sections, an eResults, 14 eTables, and 9 eFigures. Data, data, and more data, until the Supplement resembles a closet stuffed by an 8-year-old who was told to clean her room. Everything. It’s all in there.

Consider the busy physician-reader. After perusing the array of freshly published articles in the journal website’s New Online section, she may click on Hank’s title and see that long list of supplemental material populating the scholar’s margin. However transparent the author endeavored to be by providing so much information, she doesn’t have time to read it all now; she needs summaries.

AMA style advocates that “tables and figures demonstrate relationships among data and other types of information” and that “a figure should be used if the relationships are complex….Like a paragraph, each…figure should be cohesive and focused.”

With that reader in mind, the manuscript editor reformats the author’s originally supplied figures to journal style and hones each one to present the material clearly. No chartjunk, no extraneous elements, no distracting line treatments.

Flow diagrams show the numerical progression of patients through the study: the number screened for inclusion, the number excluded for these reasons, the number enrolled, and the number at each stage, with those excluded or lost to follow-up at each stage also accounted for. The last box shows how many patients made it to the end of the study or were included in the primary analysis. From top to bottom, the progression of numbers makes perfect arithmetic sense.

Figures of multiple clinical, radiologic, or histologic images are labeled to guide the reader: before surgery, 6 months after surgery, 2 years later; magnetic resonance images of brains from patients 1 and 4; or specimens from a healthy individual and a patient with disease preceding another from the patient 1 year after treatment.

Graphs are appropriate to the data presented: bars for frequencies, data markers and error bars for summary data, forest plots for meta-analyses. All axes and ticks are clearly labeled, curves are identified by direct labeling or by the inclusion of concise figure keys, and bars and data markers are a solid color for the patients who received treatment and without color for those who received placebo. The numbers of patients at risk at each time point lend additional meaning to Kaplan-Meier survival curves. Forest plots include numerical data in addition to the illustrated plot points, with labels on either side of the graph’s vertical line at 1.0 to indicate whether each data marker’s location favored treatment or no treatment.

Back to our reader. Time is short, so she starts with the abstract. Words are read quickly, their meaning filtered through her years of accumulated knowledge and absorbed. She takes in the tables next. Row upon row of data; numbers represent baseline characteristics, laboratory results, and statistical analysis. Again, the numbers are filtered for meaning and digested for information that can help the reader treat her own patients. She studies the figures, and their meaning is immediately apparent: the bar for affected patients from one age group is taller, a survival curve is higher and longer for patients who received the lower dosage, the difference between 2 clinical images before and after treatment is obvious. No filter needed. Instantly clear. Results from years of the author’s research are visually summarized, seen by the reader, grasped, and understood.

The Supplement stands ready for closer investigation, but first the point must not be obscured. State it—illustrate it—clearly.—Connie Manno, ELS

 

 

 

Have You Talked to Your Tables About the Dangers of Sex Bias?

The problem of bias is well documented in the biosciences. Even since the Health Revitalization Act of 1993, which laid out guidelines intended to ensure more equitable representation of women and minorities in federally funded scientific research, the problem persists. A 2010 study published in The Journal of Women’s Health found that, among 46 clinical studies enrolling both sexes, women comprised on average 37% of the participants, and among 69 studies, 87% did not conduct analyses by race or ethnicity, and 18% did not report differences in the racial makeup of the study sample at all. Examples of this sort abound and, setting aside the pernicious sociohistorical and nuanced biologic reasons for this phenomenon, the resulting reality is that medicine, as applied to women and minorities, is less evidence based because most research is extrapolated from a homogeneous population—white men.

But even as we attempt to resolve these problems—ensuring that guidelines are in place and that they are followed when conducting new research—there is another, more subtle way that these biases creep into the biomedical literature. Even if the study itself was conducted using a diverse population of participants, sometimes the reporting elides this fact.  As a manuscript editor I have encountered this problem more often than one would expect, and the culprit is usually the table.

In this table, as in many tables that I have encountered, “white” and “male” are the default. Women’s bodies and the bodies of racial and ethnic minorities are implied by the number of white male bodies present.

A good rule when presenting data in tables is to make sure that when you are reporting the sex of participants, if you choose to report only 1 sex, choose the sex that constitutes the majority of the sample. When reporting on racial and ethnic differences, be as specific as possible (even if these comprise a small percentage of participants). Who are the “others?”

The current edition of the AMA Manual of Style does not explicitly lay out these precautions, but in chapter 4, section 1, you will notice that every example shown for presenting data in tables follows these guidelines.

This is not merely a problem of “political correctness” or social equity—it is a question of accurate reporting and just plain good science.—Gabriel Dietz