How to write lay summaries of research articles for wider accessibility

Abstract

Journal articles are the key means for communicating scientific research. In the last century, science has become increasingly specialized such that journals commonly target researchers from ever narrower sub-disciplines. However, even in specialist journals, the research published can be relevant to scientists from other disciplines and to nonscientists including policymakers, managers, educators, and the general public (Knight 2003). Unfortunately, such broad audiences do not always find traditional articles easily accessible because they are written using an academic style that includes low readability of text and confusing jargon (Falkenberg and Tubb 2017).

An approach to enhance the accessibility of articles by broader audiences is the inclusion of “lay summaries” (hereafter referred to as summaries) alongside traditional abstracts. Summaries typically describe the issue studied in the paper, the research gap that was addressed, the key conclusion that addresses this gap written in general terms, and highlights the significance of the work with the goal of facilitating communication of the most important contribution of each manuscript across disciplines (e.g., L&O Letters https://aslopubs.onlinelibrary.wiley.com/hub/journal/23782242/about/author-guidelines, last accessed 20 December 2023; American Geophysical Union https://www.agu.org/publish-with-agu/publish/author-resources/plain-language-summary, last accessed 20 December 2023). While not replacing traditional abstracts, summaries are where authors are expected to communicate their research in less-technical ways that would appeal to new audiences (Breeze 2016). Indeed, nontechnical summaries have been advocated to increase the visibility, impact, and transparency of scientific research, particularly to nonscientific audiences (https://scientific-publishing.webshop.elsevier.com/publication-recognition/lay-summary-promote-work-outside-academia/#:~:text=Another%20reason%20to%20write%20a,academic%20background%20understand%20your%20work, last accessed 20 December 2023; Kuehne and Olden 2015). However, despite the goal of summaries, we know little about their accessibility to the target audiences, which can be defined by their readability and jargon content (Sharon and Baram-Tsabari 2014). Therefore, using this definition, we explore whether summaries are more accessible than abstracts, identify the guidance that journal publishers give to authors for writing summaries, and provide recommendations to authors, publishers, and editors to support the writing of article summaries with improved accessibility.

Despite the goal of summaries, we know little about their accessibility, which can be defined by the combination of their readability and jargon content (Sharon and Baram-Tsabari 2014). We define readability as “able to be read easily,” and jargon as the “technical terminology or characteristic idiom of a special activity or group” (Merriam-Webster, https://www.merriam-webster.com/, last accessed 20 December 2023).

There has been some previous analysis of the accessibility of summaries. In a subset of PLOS journals (PLOS Biology, Computational Biology, Genetics, Neglected Tropical Diseases, and Pathogens) readability was not higher in summaries compared to abstracts, but summaries typically had less jargon (Breeze 2016). Less jargon in summaries compared to abstracts was also found in another analysis of articles published in PLOS Computational Biology and PLOS Genetics (Rakedzon et al. 2017).

We add additional analysis to the above studies by conducting a text analysis to quantify and compare the readability and jargon of both summaries and abstracts for 1498 articles in eight ecology and general science journals that require authors to write summaries (full details in Supporting Information). Briefly, we selected eight journals related to biology, ecology, and environmental sciences that have summaries written by the authors of the original research paper with the aim of reaching a nonspecialist audience and to emphasize the key findings and broader implications of the research (Supporting Information Table S1). For each journal we then analyzed the summaries and abstracts published in 2020 and 2021. We quantified readability using the Flesh Reading Ease (FRE) index which uses the length of words and sentences to calculate the ease of reading, with scores ranging from 0 to 100 with a score of 50 or greater recommended to enable access by nonspecialists (Hartley et al. 2004; Kirkpatrick et al. 2017). We also measured jargon by applying the De-Jargonizer (Rakedzon et al. 2017) which uses a corpus of over 90 million words and identifies “jargon” words rarely encountered by a nonsubject specialist, with a level of 2–5% unfamiliar jargon proposed to be required for accurate comprehension (Rakedzon et al. 2017 and references therein). To examine what each journal views as the “ideal” summary, we also analyzed example abstracts and summaries provided to authors. Finally, the readability and jargon were analyzed using zero-inflated mixed effects models and generalized linear mixed effects models, respectively.

We found that for most journals, summaries had a low readability that was similar to that found in abstracts (Fig. 1; for another example of this pattern, see Breeze 2016). The mean readability score for both text types in all journals was around 20–25, well below 50, the threshold proposed to delineate accessibility by nonspecialists (Hartley et al. 2004; Kirkpatrick et al. 2017). This low readability is not surprising, given that guidelines provided to authors for writing summaries do not emphasize characteristics that influence readability, such as sentence length or complexity (Supporting Information Table S1). Moreover, the example texts provided in the author guidelines typically also have low readability, with only one above the accessibility threshold of 50 (scores were 8, 32, 33, 33, 39, 41, and 52).

In contrast to readability, we found that authors typically use less jargon in summaries than abstracts (significantly lower for five of the eight journals considered; Fig. 1; Breeze 2016; Rakedzon et al. 2017). This decrease of jargon in summaries likely results from specific guidelines to authors that emphasize the importance of reducing jargon (Supporting Information Table S1), and because the example texts typically had low jargon close to or under the 5% threshold proposed to facilitate accessibility.

Author guidelines provide important information to help interpret the above results. The representative eight journals we examined typically required authors to prepare relatively short summaries, most often taking the form of a single paragraph, although some are structured as bullet points. These summaries are intended to target broader audiences than the abstracts. Target audiences range from scientists from different disciplines or who were educated at the undergraduate level, to nonscientists such as decision-makers, resource managers, educators, and the general public. Finally, in all summaries, authors are requested to provide an overview of the research, with the majority also asking authors to explain the context and significance or possible implications.

Author guidelines primarily focus on reducing jargon rather than increasing readability. Six of the journal guidelines recommended removing jargon and technical language (Supporting Information Table S1), while only two journals provided guidelines on enhancing readability (e.g., shorter sentences). In addition, only three journals provided at least one (and up to six) example texts, with only one providing an example text specific to the journal (Supporting Information Table S1).

These results suggest that authors follow the instructions provided in author guidelines and that if the goal is to have summaries that are truly readable by nonspecialists, then actions must be taken to address the low readability of summaries. Next, we consider what actions should be taken to write more accessible summaries of scientific articles.

None declared.