A proposal for the inclusion of accessibility criteria in the authoring workflow of images for scientific articles

Abstract

This thesis investigates the problem of how to provide accessible images in academic articles in the research fields of Science, Technology, Engineering and Mathematics (STEM) and in particular in biomedicine. Currently, graphics in scientific articles are a critical information source and often provide essential information for a thorough understanding of scientific articles. People with visual and other impairments experience specific barriers that prevent them from accessing the information conveyed by figures. Therefore, when academic articles do not include universally accessible images, the ability to perceive and comprehend scientific information is dramatically reduced in blind readers, readers with visual impairments and other impairments (such as cognitive impairments) or readers accessing digital content under specific context and conditions (e.g., using a small and monochrome display).

The thesis reviews the current publishers’ policies and practices and the authors’ practices and attitudes in the process of image authoring and submission in STEM articles, in order to optimize the integration of accessibility requirements in the image creation and submission process. In particular, the proposal includes recommendations on how to make accessible images based on current practices in image editing and suggests the use of suitable, effective and meaningful captions and mentions as texts that can replace and describe the visual information, in order to spare authors (and publishers) the task of creating alternative texts.

The study can be divided into the following steps: • Firstly, an initial multidisciplinary literature review from the research fields of accessibility, library and information science, information visualization and semantic web is performed to assess the current proposals on the image access and description. The different disciplines contribute to define a model for the analysis and description of images and to provide technical solutions for improving their accessibility. • Secondly, an audit of the current publishing practices, policies and submission guidelines related to the accessibility of images in academic journals in the fields of biomedicine, computer science and mathematics is performed. This review assesses the current accessibility level of images and the characteristics and functions of the visual information commonly used in academic publications. • Thirdly, a survey and interviews with academic researchers in medicine and biology are performed for assessing practices and attitudes in the process of image authoring, submission and publishing. • Finally, a set of recommendations for improving the accessibility of images is proposed. It is based on the current practices of authors and publishers.

This research demonstrates a deficient accessibility of images, a lack of inclusion of accessibility issues into the current authoring workflow and proposes recommendations and guidelines on how to improve these conditions.

The results of the thesis also emphasize that the adoption of a general model of textual descriptions covering all types of images has limited results for the creation of effective text alternatives, due to the high heterogeneity of the characteristics, functions and context of the images. Hence, the thesis proposes the adoption of a model for the textual description of images based on the analysis of each type of image and adapted to each of them.

The main contribution of the proposal is that it is tailored to the specific context of STEM academic publishing and is linked to the current practices in image submission and publishing. This is the main difference of the proposal compared to the current accessibility guidelines, which are seldom general and disconnected from the policies and practices of the specific context where they should be applied. Therefore, it aims at limiting the changes demanded for the implementation of accessibility requirements in the current image publishing workflow, increasing the opportunities of an actual application of accessibility principles.

Esta tesis estudia la cuestión de cómo proporcionar imágenes accesibles en los artículos académicos de Ciencia, Tecnología, Ingeniería y Matemáticas (disciplinas “STEM” en inglés) y en particular en biomedicina. A menudo las figuras son una fuente de información esencial para la comprensión de un artículo científico. Las personas con discapacidades visuales u otras discapacidades se enfrentan a barreras que les impiden acceder a la información proporcionada por las figuras.

La tesis analiza las políticas y las prácticas actuales de los editores y las prácticas y las opiniones de los autores en el proceso de autoría y envío de imágenes científicas, con el objetivo de optimizar la integración de requerimientos de accesibilidad en el proceso de creación y sumisión de figuras en artículos científicos.

La investigación evidencia una falta de accesibilidad de las imágenes y de inclusión de pautas de accesibilidad en el actual flujo de trabajo de autoría de imágenes y propone recomendaciones y directrices para mejorar la situación.

Los resultados de la tesis además evidencian que la adopción de un modelo general de descripción textual para todos los tipos de imágenes científicas presenta limitaciones para la de creación de alternativas textuales eficaces, debido a la variedad de características, funciones y contextos de las imágenes mismas. Por ese motivo, la tesis propone la adopción de un modelo de descripción de imágenes basado en el análisis de cada tipología de imagen y adaptado a cada una de ella, de forma más específica posible.

La contribución principal de la propuesta es la de ser adaptada al contexto específico de la publicación de artículos académicos STEM y de ser conectada a las prácticas corrientes de autoría, sumisión y publicación de imágenes. La propuesta apunta a limitar los cambios en el actual flujo de trabajo editorial para la implementación de directrices de accesibilidad, aumentando las oportunidades de una verdadera aceptación de principios de accesibilidad.