|
The level of usability and accessibility of scientific journal websites is a fundamental el‐
ement of effective information transfer, and their level largely shapes the experience and
satisfaction of users. Consequently, these issues must be considered at each stage of web‐
site development, encompassing the analysis of user needs, the design and implementation
of the interface, and the subsequent testing, maintenance, and development of the website.
In recent years, there has been a growing popularity of the universal design (UD) approach, and
tools using artificial intelligence are also having an increasing impact on the website creation
process. The study described involved two separate experiments, with some participants
participating in both. In the initial phase, a comparison was conducted between the existing
website of the journal and a prototype that had been manually crafted by a designer in ac‐
cordance with Universal Design principles. In the second stage of the experiment, the same
website was compared with two versions of prototypes generated by different AI tools. The
evaluation of the websites, encompassing their usability, accessibility and impact on user
satisfaction, was conducted employing an array of research methodologies, encompassing eye
tracking, a proprietary questionnaire and automated tools for verifying compliance with WCAG
2.1 AA criteria. The results suggest that combining AI‐based generation with human‐centred
refinement provides the most effective strategy for designing scientific journal websites,
ensuring both high accessibility and a positive user experience.
|
