Article type
Abstract
Background
Scientific communication is fundamental to the research process, allowing researchers to disseminate their findings to the broader scientific community. A significant portion of this communication occurs through the publication of research manuscripts in reputable journals. Manuscripts must adhere to specific standards and conventions, with clarity and readability being of paramount importance.
For non-native English-speaking scientists, writing research manuscripts in English can be particularly challenging. This challenge often results in manuscripts that may contain grammatical errors, unclear language, and suboptimal readability, which can hinder the effective communication of research findings. Therefore, the need for tools and resources to assist non-native English-speaking scientists in improving the readability of their manuscripts is evident.
Artificial intelligence (AI) has demonstrated remarkable potential in various applications, including natural language processing and text analysis. ChatGPT, an AI-based tool, offers the potential to assist non-native English-speaking scientists in improving the readability of their manuscripts. This study aims to investigate whether ChatGPT can effectively enhance the readability of abstracts in research manuscripts authored by non-native English speakers, ultimately contributing to improved scientific communication.
Objectives
This study aims to assess the readability of abstracts in research manuscripts written by non-native English-speaking authors and to evaluate the impact of ChatGPT on the readability of these abstracts.
Methods
We performed a cross-sectional study. The sample consisted of research manuscripts published in medical journals from non-native English-speaking countries, written by non-native English-speaking authors. We randomly took 100 studies with abstracts written in English that were published between 2014 and 2023, and that were published in journals with publishers in the following countries: Argentina, Brazil, China, and Iran. For a secondary analysis, we also took 50 studies with abstracts written in English and published in USA and UK by native and non-native English-speaking countries. Then, we evaluated the readability using Grammarly before applying ChatGPT to the texts. Then, we applied ChatGPT using the prompt ` improve readability using short sentences ยด. Finally, we evaluated the readability using Grammarly score comparing the change before and after ChatGPT both in native and non-native authors.
Results and Discussion
To be presented at the colloquium.
Scientific communication is fundamental to the research process, allowing researchers to disseminate their findings to the broader scientific community. A significant portion of this communication occurs through the publication of research manuscripts in reputable journals. Manuscripts must adhere to specific standards and conventions, with clarity and readability being of paramount importance.
For non-native English-speaking scientists, writing research manuscripts in English can be particularly challenging. This challenge often results in manuscripts that may contain grammatical errors, unclear language, and suboptimal readability, which can hinder the effective communication of research findings. Therefore, the need for tools and resources to assist non-native English-speaking scientists in improving the readability of their manuscripts is evident.
Artificial intelligence (AI) has demonstrated remarkable potential in various applications, including natural language processing and text analysis. ChatGPT, an AI-based tool, offers the potential to assist non-native English-speaking scientists in improving the readability of their manuscripts. This study aims to investigate whether ChatGPT can effectively enhance the readability of abstracts in research manuscripts authored by non-native English speakers, ultimately contributing to improved scientific communication.
Objectives
This study aims to assess the readability of abstracts in research manuscripts written by non-native English-speaking authors and to evaluate the impact of ChatGPT on the readability of these abstracts.
Methods
We performed a cross-sectional study. The sample consisted of research manuscripts published in medical journals from non-native English-speaking countries, written by non-native English-speaking authors. We randomly took 100 studies with abstracts written in English that were published between 2014 and 2023, and that were published in journals with publishers in the following countries: Argentina, Brazil, China, and Iran. For a secondary analysis, we also took 50 studies with abstracts written in English and published in USA and UK by native and non-native English-speaking countries. Then, we evaluated the readability using Grammarly before applying ChatGPT to the texts. Then, we applied ChatGPT using the prompt ` improve readability using short sentences ยด. Finally, we evaluated the readability using Grammarly score comparing the change before and after ChatGPT both in native and non-native authors.
Results and Discussion
To be presented at the colloquium.