Summing Up Plain Language Summaries: A Systematic Review of Theory, Empirical Research and Guidelines

Background: Science should benefit the public, and it is argued that scientific evidence should be shared with the public in an adequately understandable manner. For non-experts and the public to understand scientific results, summaries of scientific studies written in plain language pose one proper solution. Already, a variety of plain language summary (PLS) formats have been established, some guidelines for writing PLS have been developed and empirical research on PLS has been conducted. However, terminology and research approaches as well as empirical findings in this comparatively young field are heterogeneous. To answer the question whether PLS are effective, we see a need for scrutinizing the heterogeneity and for structuring the theoretical and empirical landscape on PLS. Therefore, this review aims to synthesize current theoretical research and empirical evidence on PLS as well as PLS guidelines. The topics we deem to be most important in this context and that we focus on in this review are: PLS aims, characteristics, criteria and outcomes used to investigate them. Objective: The objective of this review is twofold: Based on current literature on PLS, the first aim is to develop a conceptual framework for the theory behind PLS; the second aim is to synthesize empirical evidence on the effectiveness of PLS. The corresponding research questions are: 1. How can we categorize and conceptualize theoretical considerations on the subject of PLS? 2. What is the empirical evidence on PLS in general and on the effectiveness of certain criteria in particular? Methods: In July 2020, we searched four databases (Web of Science, PubMed, PsycInfo, PSYNDEX) and performed additional handsearching for scientific publications and guidelines that investigate, discuss or describe PLS. PLS were defined as summaries of published scientific evidence that aim at a lay readership and that use the same communication format as the original scientific publication evidence (i.e., text). Articles were included if they empirically investigate PLS or report on the development or evaluation of PLS. Furthermore, guidelines on how to write PLS and theoretical articles referring to the four focused topics of this review (i.e., aims, characteristics, criteria and outcomes) were included. Title and abstract screenings were performed by two independent researchers. From the included articles, we extracted information separately on the four focused topics with regard to PLS. Applying a content analysis method, overarching categories for aims as well as characteristics were formulated to develop a conceptual framework which integrates PLS aims and characteristics categories as well as PLS criteria and outcomes used to investigate them. On the basis of this framework, empirical evidence on PLS was qualitatively synthesized. Results: A total of 5,481 records were identified. After removal of duplicates, 4,284 records’ titles (and, if necessary, abstracts) were screened, of which 153 articles were considered potentially relevant. These were examined in full text with regard to the inclusion criteria, leading to a final set of 72 articles that were included in the review. We found 31 theoretical, 25 empirical articles and 16 guidelines or guideline-related articles. The developed conceptual framework comprises six categories for PLS aims named Accessibility (e.g., a PLS should be easy to find), Understanding (e.g., a PLS aims to provide information that laypersons can process), Knowledge (e.g., a PLS aims to increase laypersons’ knowledge), Empowerment (e.g., a PLS should enable laypersons to make self-determined decisions), Communication of Research (e.g., a PLS aims to bridge the gap between the public and scientists), and Improvement of Research (e.g., a PLS aims to increase transparency); and six categories for PLS characteristics named Linguistic Attributes (e.g., the style of the used language), Formal Attributes (e.g., whether or not prespecified headlines are used), General Content (e.g., whether or not background information is included); Presentation of Results (e.g., the way statistical terms are reported), Presentation of Quality of Evidence (e.g., whether or not authors’ conflicts of interest are reported), and Contextual Information (e.g., information about the review process). Regarding empirical evidence on PLS, we found 13 studies that quantitatively investigated PLS using an experimental approach, three studies that qualitatively investigated PLS, and nine studies that investigated one form of PLS using an evaluative approach. Because of the studies’ heterogeneity, we synthesized the extracted information narratively based on our conceptual framework. The synthesis of the empirical studies showed that the evidence of PLS effectiveness is scarce. In these studies, different criteria that each could be linked to one of the six mentioned characteristic categories were investigated. However, the outcomes on which effectiveness was shown in these studies could only be linked to two of the aim categories, namely Accessibility and Knowledge. We thus found that the empirical evidence on PLS criteria did not cover all aim categories that were mentioned in the theoretical articles. Furthermore, it did not represent the entirety of criteria that we extracted from PLS guidelines. Conclusion and Implications: Considerable work to establish and investigate PLS formats has been done, but there is a lack of consistent terminology and conceptualization. There is scarce empirical evidence on the effect of specific PLS criteria. This evidence does not cover all theoretically mentioned aim categories and does not entirely represent criteria that are mentioned in guidelines. The conceptual framework developed in this review, as well as the systematically compiled empirical evidence on PLS effectiveness, may be of use for future guideline developers and PLS researchers.