Empirical analysis of the text structure of original research articles in medical journals.

Successful publishing of an article depends on several factors, including the structure of the main text, the so-called introduction, methods, results and discussion structure (IMRAD). The first objective of our work is to provide recent results on the number of paragraphs (pars.) per section used in articles published in major medical journals. Our second objective is the investigation of other structural elements, i.e., number of tables, figures and references and the availability of supplementary material. We analyzed data from randomly selected original articles published in years 2005, 2010 and 2015 from the journals The BMJ, The Journal of the American Medical Association, The Lancet, The New England Journal of Medicine and PLOS Medicine. Per journal and year 30 articles were investigated. Random effect meta-analyses were performed to provide pooled estimates. The effect of time was analyzed by linear mixed models. All articles followed the IMRAD structure. The number of pars. per section increased for all journals over time with 1.08 (95% confidence interval (CI): 0.70-1.46) pars. per every two years. The largest increase was observed for the methods section (0.29 pars. per year; 95% confidence interval (CI): 0.19-0.39). PLOS Medicine had the highest number of pars. The number of tables did not change, but number of figures and references increased slightly. Not only the standard IMRAD structure should be used to increase the likelihood for publication of an article but also the general layout of the target journal. Supplementary material has become standard. If no journal-specific information is available, authors should use 3/10/9/8 pars. for the introduction/methods/results/discussion sections.

Introduction

Publish and prosper is one of the sayings scientists often encounter. Working in the field of research means constant publishing. The competition among scientists is strong, and journal space is limited. However, the world of publication can be a black box, and writing is challenging for many [1]. Concurrently, the art of scientific communication is rarely taught, and scientific writing distinguishes fundamentally from literary writing. Only a few authors focus on the process of writing. One such procedural system was developed by Albert [2], and he demystified article writing using a 10-step process [3]. In one of these steps, Albert’s concept looks at a sales model for the article. One key element of this sales approach to publication is that the target audience is the journal editor because the editor is the gatekeeper for acceptance or rejection of an article or even the early rejection, sometimes called desk-rejections. As always the first impression of a manuscript is the best impression [4]. Thrower has provided reasons why he accepted [5] or rejected [6] manuscripts for publication. One important aspect is the presentation of the material [7]. It is clear that authors should follow author instructions of the target journal. The referencing pattern also plays a role, and “a well formatted paper makes the editor happy as he need not to do anything further from his side as far as presentation is concerned” [4]. Reasons for early rejections of articles have been given in an editorial by Froese and Bader [8], and they stated that manuscripts not following the structure a standard article are likely to get rejected. In this context, we need to look at the question what the typical structure of an article is.

The typical structure for the main text of an article in a medical journal is the introduction, methods, results and discussion (IMRAD) structure. Although it was recommended as standard structure at the beginning of the 20th century, it became the primary structure only after 1965, and it was the only article structure in the 1980s [9]. The major change in the mid 1960s is most likely related to a conference of medical editors held at the 19th General Assembly of the World Medical Association [10], during which Hill [11] suggested that research articles should answer the four questions: why did you start, what did you do, what answer did you get and what does it mean anyway. Little research has been done on details of these four sections of an article [2, 12]. Albert [2] an analysis of 50 consecutive articles published after June 1, 1997, from each of the 6 journals Archives of Disease in Childhood (Arch Dis Child), Journal of Pediatrics (J Pediatr), Pediatric Research (Pediatr Res), The BMJ (BMJ), The Lancet (Lancet) and The New England Journal of Medicine (NEJM). He reported the means and standard deviations (SD) of the number of paragraphs per section as given in Table 1. From this, Albert [2] derived the `typical journal structure`, which he calls 2/7/7/6 and which means that 2 paragraphs should be used for the introduction, 7 for methods, 7 for results and 6 for the discussion. Albert [3] was less rigorous and suggested a 2-3/4-6/4-6/5-8 structure. Other recommendations based on Albert’s empirical work have also been proposed, such as 2/5/5/4 for shorter articles [13]. Soares de Araújo [12] considered original articles published in the January 2012 and 2013 issues of Arquivos Brasileiros de Cardiologia and the first two issues of the Journal of the American College of Cardiology from the same years and recommended 3/6-9/4-9/≤10 paragraphs. If we meta-analyze the data provided by Albert [2] per section using the DerSimonian and Laird [14] approach, the recommended structure is 3/8/7/7 (Table 1), i.e., 25 paragraphs in total. Albert’s work is, however, more than 20 years old and has not been updated.

Table 1

Means and standard deviations for number of paragraphs per section.

Modified from Albert [2], Fig 5.1.

Journal	Introduction	Methods	Results	Discussion	Structure
Arch Dis Child	2.7 ± 1.3	6.5 ± 4.0	6.1 ± 4.0	6.9 ± 2.8	3/7/6/7
BMJ	2.3 ± 0.9	6.0 ± 3.7	5.9 ± 3.1	7.4 ± 2.8	3/6/6/7
J Pediatr	2.6 ± 1.1	6.7 ± 3.4	7.0 ± 3.9	7.3 ± 2.8	3/7/7/7
Lancet	2.6 ± 1.3	7.6 ± 3.6	6.1 ± 2.9	7.0 ± 2.6	3/8/6/7
NEJM	2.6 ± 1.1	9.2 ± 3.3	8.9 ± 3.8	6.9 ± 1.8	3/9/9/7
Pediatr Res	3.0 ± 1.3	9.6 ± 3.8	6.3 ± 2.9	8.5 ± 3.4	3/10/6/9
Meta-analysis	2.6 ± 0.5	7.6 ± 1.5	6.6 ± 1.4	7.2 ± 1.0	2-3/6-9/6-7/6-8

Arch Dis Child: Archives of Diseases in Childhood, BMJ: The BMJ, J Pediatr: Journal of Pediatrics, NEJM: The New England Journal of Medicine, Pediatr Res: Pediatric Research. Meta-analysis: results from DerSimonian and Laird [14] approach.

The first aim of our article is to provide recent results on characteristics of the structure of original articles published in major medical journals. Since the reporting of studies changed over the past 20 years due to the availability of reporting guidelines, such as the CONSORT [15] or STARD [16] statement, we hypothesize that more recent journal articles have an increasing number of paragraphs over time, especially in the methods section. To this end, we analyze data from randomly selected original articles published in years 2005, 2010 and 2015 from the journals BMJ, The Journal of the American Medical Association (JAMA), Lancet, NEJM and PLOS Medicine (PLOS). We expect that PLOS, an electronic only journal, has more paragraphs than the print journals BMJ, JAMA, Lancet and NEJM because there are, in principle, no page restrictions. The second aim of our work is to expand the statistics to other structural elements, i.e., the number of tables, figures and references as well as the availability of supplementary material. We expect that recently published articles have more likely supplementary material.

Materials and methods

In original articles published in the English language medical journals BMJ, JAMA, Lancet, NEJM and PLOS in the years 2005, 2010 and 2015, we investigated the number of paragraphs per section. Per journal and year of publication we randomly selected 30 articles, totaling to 450 original research articles. We checked the presence of the IMRAD structure, counted the number of tables, figures and references and checked if supplementary material was available for an article. Data were extracted by M.R.

For each year and each journal, means and SD were calculated for continuous outcomes, and absolute and relative frequencies were reported for the availability of supplementary material. The DerSimonian and Laird [14] approach was used to perform random effect (RE) meta-analyses which allows for variability in the number of paragraphs between journals and over time. These analyses were performed within each journal over the years, within each year over the journals, and for all years and all journals together. Pooled RE estimates and standard errors were calculated. The effect of time was investigated by linear mixed models with journal as RE and year as fixed effect (FE) for the number of paragraphs and the proportion of journals with supplementary information. Effect estimates and corresponding 95% CIs were reported. The specific hypothesis that PLOS has more paragraphs than the print journals was investigated with a mixed linear model with year as RE and journal as FE. Methodological details are provided in the S1 Appendix.

No adjustments were made for multiple testing, and the significance level was set to 0.05 for all analyses. All statistical analyses were done in R version 3.6.2. Data and analysis code in Markdown are available as supplementary material.

Results

A total of 450 original research articles from 5 medical journals were analyzed. The IMRAD structure was used in all of them. S1 Table displays the descriptive statistics for the number of paragraphs per section for each year and each journal. An increase in the number of paragraphs per section and in total can be seen for all journals over time. The increase is, however, small for JAMA and NEJM. An obvious increase in the methods section can be observed for BMJ and Lancet with an average of 5 and four additional paragraphs, respectively (Fig 1). The increase in the number of paragraphs was also visible in PLOS with one, 5, three and two additional paragraphs in the introduction, methods, results and discussion for years 2015 when compared with 2005 (S1 Table and Fig 1).

Fig 1

Forest plots from meta-analyses.

Pooled means of number of paragraphs (black squares) and 95% confidence intervals (CI, lines) are displayed for each journal for A) methods section in 2005, B) methods section in 2010, C) methods section in 2015 and D) the total number of paragraphs in 2015, respectively. Summary statistics (black diamond) was calculated using the random effects DerSimonian & Laird approach [14]. BMJ: The BMJ, JAMA: The Journal of the American Medical Association, NEJM: The New England Journal of Medicine, PLOS: PLOS Medicine.

While the number of figures and tables overall remained rather constant, the number of references increased between 2005 and 2010 and 2015, respectively, with the largest increase in the number of references for BMJ and PLOS (S1 Table). The most striking change is seen for the number of original articles with supplementary material. For example, while no JAMA article had supplementary material in 2005, 97% (29 of 30) had supplementary material in 2015.

Table 2 shows article characteristics by year averaged over the 5 journals. The smallest and largest increase per year for the number of paragraphs was observed for the introduction (0.03 pars. per year, 95% CI: 0.00–0.06) and methods sections (0.29 pars. per year, 95% CI: 0.19–0.39). While the standard article structure was 3/9/9/8 in 2005, it increased to 3/11/10/8 in 2015. On average within 10 years, methods thus increased by two paragraphs and results by one paragraph. This corresponds to an increase of approximately one paragraph every two years (1.08 pars. per 2 years; 95% CI: 0.70–1.45). Consideration of all years and all journals, the pooled standard article structure is 3/10/9/8.

Table 2

Results of meta-analyses per year over journal.

Year	Introduction	Methods	Results	Discussion	Total	Tables	Figures	References	Suppl
2005	3.19 ± 0.53	9.27 ± 1.59	8.61 ± 1.49	7.72 ± 0.94	28.55 ± 2.68	3.27 ± 0.72	1.87 ± 0.62	30.65 ± 4.77	0.34 ± 0.13
2010	3.08 ± 0.49	10.85 ± 1.64	8.84 ± 1.70	8.19 ± 1.23	31.41 ± 3.01	3.18 ± 0.66	2.45 ± 0.77	35.08 ± 5.60	0.73 ± 0.09
2015	3.36 ± 0.45	11.36 ± 1.89	9.61 ± 1.62	8.29 ± 1.26	32.76 ± 3.49	2.90 ± 0.57	2.55 ± 0.67	34.34 ± 6.25	0.97 ± 0.02
All years	3.22 ± 0.28	10.39 ± 0.98	9.00 ± 0.92	8.00 ± 0.64	30.54 ± 1.74	3.08 ± 0.37	2.26 ± 0.39	32.97 ± 3.14	0.67 ± 0.11
Change per year	0.03	0.29	0.14	0.08	0.54	-0.02	0.08	0.75	0.06
95% CI	0.00–0.06	0.19–0.39	0.04–0.23	0.01–0.16	0.35–0.73	-0.06–0.02	0.03–0.13	0.30–1.20	0.05–0.07
p-value	0.050	< 0.001	0.004	0.036	< 0.001	0.336	0.001	0.001	< 0.001

Pooled means and respective standard errors for the number of paragraphs per section, the total number of paragraphs (Total), the number of tables, figures and references from DerSimonian and Laird [14] meta-analysis per year over journals. The proportion of articles with supplementary material (Suppl) is displayed in the last column. The change in the number of paragraphs per year/percent of articles with supplementary material per year from a linear mixed model is displayed together with its 95% confidence interval (95% CI) and p-value.

The number of tables did not alter (-0.02 per year; 95% CI: -0.06–0.02), and figures increased slightly (0.08 per year; 95% CI: 0.03–0.13). Per year 0.75 more references were observed when averaged over the journals (95% CI: 0.30–1.20). This increase is, however, primarily caused by BMJ and PLOS, while the number of references was similar over the years for JAMA, Lancet and NEJM.

During this period, the percentage of articles with supplementary material increased from 39% in 2005 to 93% in 2015. This confirms our hypothesis about the increasing number of supplementary material per years (p < 0.001).

Results per journal pooled over the years are displayed in Table 3 and S1 and S2 Figs. The only considered online only journal PLOS had the largest number of paragraphs in the introduction, results and discussion sections. While PLOS did not have significantly more paragraphs in the methods section, it had approximately one additional paragraph in the introduction and the discussion, respectively. On average, PLOS articles hat 3.52 paragraphs more than the other journals (95% CI: 1.60–5.43). The NEJM had the lowest number of paragraphs per article among the 5 journals considered. It is the only journal still with a 2 before the decimal point for the average number of paragraphs in the introduction. Only BMJ and NEJM had less than 10 paragraphs on average for the methods, and NEJM had approximately three fewer paragraphs in the discussion compared with the other four medical journals considered. Articles published in NEJM also had the lowest number of tables and references. However, BMJ and Lancet had the lowest number of figures. Unexpectedly, the online only journal PLOS did not have more tables than the print journals (0.10; 95% CI: -0.29–0.49), but 1.63 (95% CI 1.14–2.13) more figures and 14.45 (95% CI: 9.85–19.05) more references compared to the print journals.

Table 3

Results of meta-analyses per journal over years.

Journal	Introduction	Methods	Results	Discussion	Total	Tables	Figures	References
BMJ	3.43 ± 0.56	9.50 ± 2.38	8.78 ± 2.67	8.80 ± 1.91	28.91 ± 3.75	3.36 ± 1.06	1.80 ± 0.94	31.23 ± 9.10
JAMA	3.12 ± 0.57	11.21 ± 1.96	8.94 ± 1.85	9.30 ± 1.47	32.44 ± 3.49	3.48 ± 0.75	1.83 ± 0.78	35.18 ± 7.33
Lancet	3.01 ± 0.63	11.06 ± 2.29	8.79 ± 2.22	9.20 ± 2.22	31.93 ± 5.20	2.98 ± 0.84	2.82 ± 0.92	32.27 ± 6.89
NEJM	2.94 ± 0.60	9.67 ± 1.82	9.08 ± 1.56	6.88 ± 0.88	28.57 ± 3.02	2.66 ± 0.65	2.29 ± 0.67	30.40 ± 4.87
PLOS	4.05 ± 0.93	10.71 ± 2.98	9.45 ± 2.79	9.81 ± 2.25	34.48 ± 5.62	3.38 ± 1.20	3.95 ± 1.87	46.84 ± 11.58
PLOS vs. others	1.05	0.70	0.86	0.91	3.52	0.10	1.63	14.45
95% CI	0.76–1.33	-0.30–1.69	-0.09–1.82	0.11–0.1.71	1.60–5.43	-0.29–0.49	1.14–2.13	9.85–19.05
p-value	< 0.001	0.200	0.080	0.030	< 0.001	0.60	< 0.001	< 0.001

Pooled means and respective standard errors for the number of paragraphs per section, the total number of paragraphs from DerSimonian and Laird [14] meta-analysis per journal over years. The number of additional paragraphs of PLOS Medicine compared with the other four journals as estimated by a linear mixed model is displayed together with the corresponding 95% confidence interval (95% CI) and p-value. BMJ: The BMJ, JAMA: Journal of the American Medical Association, NEJM: The New England Journal of Medicine, PLOS: PLOS Medicine.

Discussion

All original articles from the 5 major medical journals considered followed the IMRAD structure. From 2005 to 2015, the total number of paragraphs increased by one every two years, and the largest increase was observed for the methods section. While the NEJM had a large number of paragraphs in the Albert analysis from 1996 [2], it had the smallest number of paragraphs averaged over the years. As expected, PLOS had the highest number of paragraphs, and PLOS articles were approximately 3.5 paragraphs longer than articles in the print journals. PLOS also had the largest number of figures and references per article. Compared with 2005, it is now standard for all 5 investigated journals to have supplementary material.

Albert used the number of paragraphs instead of words as measure for the text structure of a scientific article because paragraphs are a more manageable unit than words alone [2]. Paragraphs keep the potential reader interested when they are written so that they “become in effect a series of inverted triangles”. This means that the first sentence of a paragraph is the key sentence or topic sentence. The following sentences should be only supportive and elaborate the topic sentence.

In addition, paragraphs are related to items from publication recommendations. The EQUATOR network–EQUATOR stands for Enhancing the QUAlity and Research Transparency Of health Research–provides reporting guidelines and checklists for a wide variety of research and study types as help for authors to make their research transparent. Several journals require authors to fully adhere to these guidelines. Some of the guidelines have been updated over time due to practical experience. For example, the first version of the CONSORT (Consolidated Standard of Reporting Trials) statement was published in 1996 and included 21 items. The last revision, termed CONSORT 2010 provides a checklist with 25 items, of which 12 are further divided into a and b, making 37 items in total. Similarly, the STARD (Standards for Reporting of Diagnostic Accuracy Studies) statement for the reporting of diagnostic studies has also increased in the number of items from its first version, which was published in 2004, to its current edition from 2015. While the first STARD statement has 25 items, the STARD 2015 statement has 30 items, 4 of which have an a/b division, making 34 in total. An increase in the number of paragraphs can therefore be expected if a publication recommendation is updated and the number of items increased.

Based on his empirical work from almost 25 years back in time, Albert recommended the use of 2/7/7/6 paragraphs for the introduction/methods/results/discussion sections. Other authors were less strict in their recommendations [3, 12]. We provide an update about the standard structure for an article, and we now generally recommend 3/10/9/8 paragraphs for the four main sections. The total number of paragraphs thus is approximately 30. This directly leads to a recommendation for the number of words per paragraph. For example, JAMA has a word limit of 3000. With 30 paragraphs this gives approximately 100 words per paragraph in the main text. Soares de Araújo [12] recommended 130 words for cardiovascular journals because these have an upper word count limit of approximately 4000 words after subtraction of references.

Suggestions for the topics of the different paragraphs have been provided by Albert [2] and Soares de Araújo [12]. Their suggestions for topics differ, however, substantially. Specifically, the first paragraph of the discussion is a brief summary of the main findings according to Albert, while the study problem should be discussed again in the first paragraph of the discussion according to Soares de Araújo. In our opinion, the repetition of the study problem is superfluous, and we agree with Albert’s general concept of article writing. He suggested 6 topics for the discussion section: 1) summary of main findings, 2) weaknesses of the study, 3) strengths of the study, 4) how it fits in the literature, 5) implications for future research and 6) implications for policy/treatment. If 8 sections are used in the discussion, the summary of the main findings and its fit to the literature are generally expanded by one additional paragraph each, but the topics are not changed per se. Instead, it seems very natural that the fit in the published literature is more elaborated, e.g., by integrating related topics, systematic reviews and judging original studies, such as randomized controlled trials. Writers and scientists also presented concepts how the complex problem of writing a whole journal article can be divided into smaller problems. One such approach was suggested by O’Connor and Holmquist [17], and another concept was developed by Albert [2]. The ideas presented differ substantially, and this is best illustrated from the “summary statements” [17] or the “message” [2]. Albert does not get weary of emphasizing the importance of the message of the paper. The message should have 12 words, should have a verb, should not be a question and, in our experience, should not include an “and”. The paper is then streamlined along this single message. This is in contrast to the concept of O’Connor and Holmquist [17] because a manuscript cannot be written according to a single message if there are up to three “conclusions summarizing the major contributions of the manuscript to the scientific community”.

In our study, we did not investigate the effect of the article structure on citation frequency. Several studies observed relationships between the length of the title and citation frequency [18-24]. Shorter titles have a higher citation frequency, and the conclusion is: keep the title short. Linguistic complexity of title, abstract and main text has also been studied in relation with citation frequency [20, 25–28]. While no difference was found between citation frequency and linguistic complexity of the main text [28], top ranked journals use a simple language in title and abstract [27]. However, scientific articles are generally difficult to read [29, 30]. For example, the Gunning fog index, which looks at sentence length and word complexity [31], is approximately 17 for scientific articles [32-34]. Although text complexity is reduced are after peer-review, texts are still substantially more complex compared to daily newspaper articles, which have a fog index of 12 [35]. Insurance policies are in contrast even more complex with a fog index of approximately 20 [29].

Authors also investigated the association between citation frequency, page length, the number of references and author recognition [36-38]. It is obvious that larger articles may represent review articles, as may articles with a higher number of references. With these arguments, it can already be expected that articles with more pages and more references have higher citation numbers.

A reviewer of our work has pointed to the importance of investigating the influence of media paying possibly more attention to journal articles in the past years and changes in journal strategies and instructions for authors. These aspects should be assessed in future research.

In conclusion, authors should not only use the standard IMRAD structure to increase the likelihood for publication of their work. They should also take into account the general layout of their target journal. If a journal-specific structure is not at hand, authors should use 3, 10, 9 and 8 paragraphs for the introduction, methods, results and discussion sections, respectively. Supplementary material has become a standard and should be used when deemed appropriate. Authors should be aware that print journals might differ in their structure from online only journals because of the absence of page limits for online articles. Finally, and most importantly, the instructions to authors of the target journal must definitely be met.

Data.

(XLSX)

Click here for additional data file.

R Markdown file for analyses.

(RMD)

Click here for additional data file.

Forest plots from meta-analyses for introduction, results and discussion section.

Pooled means of number of paragraphs (black squares) and 95% confidence intervals (CI, lines) are displayed for each journal for A) introduction section in 2005, B) introduction section in 2015, c) results section in 2005, D) results section in 2015, E) discussion section in 2005 and F) discussion section in 2015, respectively. Summary statistics (black diamond) was calculated using the random effects DerSimonian & Laird approach (14). BMJ: The BMJ, JAMA: Journal of the American Medical Association, NEJM: The New England Journal of Medicine, PLOS: PLOS Medicine.

(EPS)

Click here for additional data file.

Forest plots from meta-analyses for tables, figures and references.

Pooled means of number of tables, figures and references (black squares) and 95% confidence intervals (CI, lines) are displayed for each journal for A) tables in 2005, B) tables in 2015, c) figures in 2005, D) figures in 2015, E) references in 2005 and F) references in 2015, respectively. Summary statistics (black diamond) was calculated using the random effects DerSimonian & Laird approach (14). BMJ: The BMJ, JAMA: Journal of the American Medical Association, NEJM: The New England Journal of Medicine, PLOS: PLOS Medicine.

(EPS)

Click here for additional data file.

Descriptive statistics per journal and per year.

Means and standard deviations for the number of paragraphs per section, the total number of paragraphs (Total), the number of tables, figures and references by journal and year of publication. The last column provides absolute frequencies and relative frequencies (in parenthesis) for the availability of supplementary material (Suppl). BMJ: The BMJ, JAMA: The Journal of the American Medical Association, NEJM: The New England Journal of Medicine, PLOS: PLOS Medicine.

(DOCX)

Click here for additional data file.

(DOCX)

Click here for additional data file.

15 Jul 2020

PONE-D-20-15797

Empirical Analysis of the Text Structure of Original Research Articles in Medical Journals

PLOS ONE

Dear Dr. Ziegler,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I would like authors to clarify the importance of word limits rather than number of paragraphs or they can clarify why it is important for them in discussion. I would like authors to add discussions on the reasons for changes in publication behavior of authors as well.

Please submit your revised manuscript by Aug 29 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Omid Beiki, M.D., Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Financial Disclosure section:

'The author(s) received no specific funding for this work.'

We note that one or more of the authors are employed by a commercial company: Dr. Hüsing Aktuar GmbH

a. Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

b. Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc.

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests) . If this adherence statement is not accurate and  there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

c. Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

3. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: N/A

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Comments of Authors

General Comments

Effective presentation of methods, findings, and importance are of critical importance in obtaining a positive response from journal editors. Important work may not be published if the methods, results, and implications are not stated clearly. My main concern is that the paper is descriptive, but does not deal with why patterns are as they are and why these patterns have evolved. The changes may reflect an increased sophistication of the readership, especially about methods. Large data sets are now more common, and these require more detailed description which reflects both the methods sections and presentation of supplementary material. The increase in supplementary material probably also reflects changes in computer technology, e.g., the cloud. Have the instructions to the authors changed? Has demand for journal space increased, especially among the top-ranked journals? Is the media paying more attention to journal articles than in the past which would affect how papers are written?

Specific Comments

Lines 65 ff. A more compelling justification for focusing on paragraphs is needed. Years ago, I was taught that among the sections, relatively few readers focus on the methods section. This is why this section is sometimes in small print. This section seems to be gaining in importance, but I do not see the link to the number of paragraphs as the best measure of this.

Lines 86-87. The increase in number of paragraphs in the methods section may reflect increasing sophistication of the readership, especially about statistical methods.

Line 106. What is the rationale for using the random effects method? Why is it preferred over the other statistical methods?

Line 120. The paper does not adequately document changes in editorial policy. Have instructions to the authors changed? Changes in word limits and the extent to which they are enforced? This comment also applies elsewhere? It might have been a good idea to have interviewed journal editors to document changes in their policies.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

25 Aug 2020

We are very grateful to both reviewers for their feedback and helpful comments which have led to considerable improvement of our manuscript. Reviewer comments are highlighted in blue, and answers to their comments are provided in black in this document. Added and/or altered text is given in italic. These changes are highlighted in the revised manuscript. In addition, we provide a clean version of our manuscript.

Review Academic Editor

I would like authors to clarify the importance of word limits rather than number of paragraphs or they can clarify why it is important for them in discussion.

We thank the reviewer for the comment and the opportunity to provide an additional justification for the choice of the number of paragraphs as measure in our work. The Introduction, Methods, Results and Discussion (IMRAD) structure can be traced back to Hill (1). His work in the mid 1960 was motivated by inadequate reporting of journal articles. More recently, recommendations for publications have been published for various study designs by the EQUATOR (Enhancing the QUAlity and Research Transparency Of health Research) network (www.equator-network.org). This network does not only provide reporting guidelines but also checklists for a wide variety of research and study types as help for authors to make their research transparent. Practical experience with guidelines has led to revisions. For example, the first version of the CONSORT (Consolidated Standard of Reporting Trials) statement was published in 1996 (2), revised 5 years later (3) and again in 2010, leading to the current CONSORT 2010 statement (4). While the CONSORT statement from 1996 only included 21 items, CONSORT 2010 provides a checklist with 25 items, of which 12 are further divided into a and b, making 37 items in total.

The STARD (Standard for Reporting of Diagnostic Accuracy Studies) statement for the reporting of diagnostic studies has also increased in the number of items. The first statement from 2004 had 25 items, while the revised version from 2015 has 30 items, 4 of which have an a/b division, making 34 in total.

In our opinion, items are closely related to paragraphs, not to the number of words.

Even more importantly, Albert (5) described that authors should look at articles recently published in their target journal and count the number of paragraphs. He stated the number of paragraphs is “a more manageable unit than words alone”. Furthermore, paragraphs keep the potential reader interested when paragraphs are written as a series of inverted triangles (5). The first sentence of a paragraph is the key sentence or topic sentence. The following sentences are only supportive and elaborate the key sentence.

Overall, we are therefore convinced that paragraph is the concept which is simpler to handle, reflects publication recommendations and an important scientific information because of the topic sentences.

We have integrated these arguments into the manuscript, and the added discussion reads:

Albert used the number of paragraphs instead of words as measure for the text structure of a scientific article because paragraphs are a more manageable unit than words alone (2). Paragraphs keep the potential reader interested when they are written so that they “become in effect a series of inverted triangles”. This means that the first sentence of a paragraph is the key sentence or topic sentence. The following sentences should be only supportive and elaborate the topic sentence.

In addition, paragraphs are related to items from publication recommendations. The EQUATOR network – EQUATOR stands for Enhancing the QUAlity and Research Transparency Of health Research – provides reporting guidelines and checklists for a wide variety of research and study types as help for authors to make their research transparent. Several journals require authors to fully adhere to these guidelines. Some of the guidelines have been updated over time due to practical experience. For example, the first version of the CONSORT (Consolidated Standard of Reporting Trials) statement was published in 1996 and included 21 items. The last revision, termed CONSORT 2010 provides a checklist with 25 items, of which 12 are further divided into a and b, making 37 items in total. Similarly, the STARD (Standards for Reporting of Diagnostic Accuracy Studies) statement for the reporting of diagnostic studies has also increased in the number of items from its first version, which was published in 2004, to its current edition from 2015.While the first STARD statement has 25 items, the STARD 2015 statement has 30 items, 4 of which have an a/b division, making 34 in total. An increase in the number of paragraphs can therefore be expected if a publication recommendation is updated and the number of items increased.

I would like authors to add discussions on the reasons for changes in publication behavior of authors as well.

We agree with the reviewer that the publication behavior of authors changed in the last years as well. Thousands of articles are submitted to scientific journals each year. For example, PLOS ONE states on its website (https://journals.plos.org/plosone/s/journal-information, access: August, 10th 2020) that “more than 20,000 new authors are joining the PLOS ONE community every year in over 200 research areas”. Competition among scientists is strong, and there is only limited space for publications. Authors thus become more and more aware of the importance of writing their articles according to the available publication recommendations and instructions of authors to increase their chances of a successful publication.

We have altered the introduction which now reads:

Publish and prosper is one of the sayings scientists often encounter. Working in the field of research means constant publishing. The competition among scientists is strong, and journal space is limited. However, the world of publication can be a black box, and writing is challenging for many (1). Concurrently, the art of scientific communication is rarely taught, and scientific writing distinguishes fundamentally from literary writing. Only a few authors focus on the process of writing. […]

Reviewer 1

First of all, we want to thank Reviewer 1 for the useful suggestions. We fully agree that several suggestions, such as the influence of media, change of the instructions for authors and interviewing editors are of importance. However, these suggestions are beyond the scope of our own work, and they should be addressed in future research. We added the suggestions in the discussion, and we provide further information below.

General Comments

Effective presentation of methods, findings, and importance are of critical importance in obtaining a positive response from journal editors. Important work may not be published if the methods, results, and implications are not stated clearly. My main concern is that the paper is descriptive, but does not deal with why patterns are as they are and why these patterns have evolved. The changes may reflect an increased sophistication of the readership, especially about methods. Large data sets are now more common, and these require more detailed description which reflects both the methods sections and presentation of supplementary material.

In our reply to the first comment from the Academic Editor, we have hopefully addressed this aspect in sufficient detail.

Furthermore, we focus again to the CONSORT statement: CONSORT 2010 (4) provides a minimum set of recommended items for reporting a randomized clinical trial which improves the quality of adequate reporting of a trial. Due to that, readers may better understand specific aspects of the trial, such as its design, conduct, analysis or interpretation of the results. Even more, readers may assess the validity of the trial results. Publication recommendations thus facilitate complete and transparent reporting. In our opinion, this increases the chance of a publication in the target journal or may even be a prerequisite. In 1996, CONSORT was a 21-item checklist (2). The revised CONSORT 2001 statement included a 22-item checklist (3). In greater detail, items that were previously combined were now separated, e.g., outcomes (item 6) and sample size (item 7) were separated. Moreover, some items requested additional information.

In 2010, three more items were added. This reflects the need of more information provided by authors.

In summary, reporting guidelines use items to structure the scientific work, and in our opinion items reflect paragraphs.

We have added these arguments to the discussion which now reads:

Albert used the number of paragraphs instead of words as measure for the text structure of a scientific article because paragraphs are a more manageable unit than words alone (2). Paragraphs keep the potential reader interested when they are written so that they “become in effect a series of inverted triangles”. This means that the first sentence of a paragraph is the key sentence or topic sentence. The following sentences should be only supportive and elaborate the topic sentence.

In addition, paragraphs are related to items from publication recommendations. The EQUATOR network – EQUATOR stands for Enhancing the QUAlity and Research Transparency Of health Research – provides reporting guidelines and checklists for a wide variety of research and study types as help for authors to make their research transparent. Several journals require authors to fully adhere to these guidelines. Some of the guidelines have been updated over time due to practical experience. For example, the first version of the CONSORT (Consolidated Standard of Reporting Trials) statement was published in 1996 and included 21 items. The last revision, termed CONSORT 2010 provides a checklist with 25 items, of which 12 are further divided into a and b, making 37 items in total. Similarly, the STARD (Standards for Reporting of Diagnostic Accuracy Studies) statement for the reporting of diagnostic studies has also increased in the number of items from its first version, which was published in 2004, to its current edition from 2015.While the first STARD statement has 25 items, the STARD 2015 statement has 30 items, 4 of which have an a/b division, making 34 in total. An increase in the number of paragraphs can therefore be expected if a publication recommendation is updated and the number of items increased.

The increase in supplementary material probably also reflects changes in computer technology, e.g., the cloud.

It could well be that we have misunderstood the reviewer, and we partly agree with the reviewer. In our opinion, there are two main reasons for the increase of supplementary material. First, the research itself and the reporting of research is getting more complex. Study data sets are getting larger, and the increased data sets need more detailed and longer descriptions, especially in the Methods and the Results sections. Results from primary or main analyses are generally presented, while sensitivity analyses and additional findings are in the supplement. This allows to the keep focus on the main aspects and findings of the article. Second, and this might be the argument of the reviewer, electronic publishing permits longer articles because, in principle, there are no page restrictions. The possibility of electronic publications (pdf articles in addition to printed articles) provides the space to place further information in supplementary material.

Have the instructions to the authors changed? Has demand for journal space increased, especially among the top-ranked journals? Is the media paying more attention to journal articles than in the past which would affect how papers are written?

We fully agree with the reviewer that this aspect is of interest. However, it is beyond the scope of our work and should be left for future research. We have added this point in the discussion of our revised manuscript.

The demand for journal space has substantially increased. This is well reflected by the number of journals which increased over time.

For example, the number of journals listed in

Index Medicus and MEDLINE (https://www.nlm.nih.gov/bsd/index_stats_comp.html) increased substantially over time. Furthermore, there are more authors per year (see, e.g., for the USA from 2011 to 2019: https://www.statista.com/statistics/572476/number-writers-authors-usa/). Space in high ranked journals has not increased accordingly.

Media might be important as well. However, the role of media is beyond the scope of our work. We have added this aspect in the discussion, and the text now reads:

A reviewer of our work has pointed to the importance of investigating the influence of media paying possibly more attention to journal articles in the past years and changes in journal strategies and instructions for authors. These aspects should be assessed in future research.

\f

Specific Comments

Lines 65 ff. A more compelling justification for focusing on paragraphs is needed. Years ago, I was taught that among the sections, relatively few readers focus on the methods section. This is why this section is sometimes in small print. This section seems to be gaining in importance, but I do not see the link to the number of paragraphs as the best measure of this.

Above, we have argued that the number of paragraphs is naturally related to the publication recommendations and revised our text in the discussion.

Lines 86-87. The increase in number of paragraphs in the methods section may reflect increasing sophistication of the readership, especially about statistical methods.

In our opinion, the increase of the number of paragraphs, particularly in the method section, is reflected by the available publication recommendations. We have incorporated the arguments in the Discussion.

Line 106. What is the rationale for using the random effects method? Why is it preferred over the other statistical methods?

We chose the random effect (RE) model because this meta-analysis model permits the true effect sizes, i.e., the number of paragraphs of a section, to vary between the journals and the years. This heterogeneity is taken into account in a meta-analytical model with RE. In contrast, the meta-analytical model with fixed effects (FE) assumes that the number of paragraphs of a section is always the same for all journals and at all time points. Only random fluctuation would be permitted in this model. However, this assumption does not correspond with our assumptions that 1. the number of paragraphs of a section varies between the considered medical journals and 2. the number of paragraphs of a section varies over time.

We have rephrased the methods as follows:

The DerSimonian and Laird (14) approach was used to perform random effect (RE) meta-analyses which allows for variability in the number of paragraphs between journals and over time.

Line 120. The paper does not adequately document changes in editorial policy. Have instructions to the authors changed? Changes in word limits and the extent to which they are enforced? This comment also applies elsewhere? It might have been a good idea to have interviewed journal editors to document changes in their policies.

We fully agree with the reviewer that changes in the editorial policy might have played a role. However, the change of the instructions of authors is beyond the scope of our work.

We have rewritten the discussion which now reads:

A reviewer of our work has pointed to the importance of investigating the influence of media paying possibly more attention to journal articles in the past years and changes in journal strategies and instructions for authors. These aspects should be assessed in future research.

References

(1) Hill AB. The reasons for writing. BMJ. 1965; 2:870.

(2) Begg C et al. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA. 1996; 276:637-9

(3) Moher D, Schulz KF, Altman D; CONSORT Group (Consolidated Standards of Reporting Trials). The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA. 2001; 285:1987-91

(4) Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 Statement: Updated Guidelines for Reporting Parallel Group Randomised Trials. PLOS Med 2010;7: e1000251.

(5) Albert T. Winning the Publications Game: The Smart Way to Write Your Paper and Get It Published. 4th ed. Boca Raton: CRC Press; 2016.

Submitted filename: Response_to_Reviewers.docx

Click here for additional data file.

24 Sep 2020

Empirical Analysis of the Text Structure of Original Research Articles in Medical Journals

PONE-D-20-15797R1

Dear Dr. Ziegler,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Omid Beiki, M.D., Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

28 Sep 2020

PONE-D-20-15797R1

Empirical Analysis of the Text Structure of Original Research Articles in Medical Journals

Dear Dr. Ziegler:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Omid Beiki

Academic Editor

PLOS ONE