Abstract:

In recent years, research in the field of smart learning environments (SLE) has dealt extensively with various technical possibilities that could realize such spaces. According to Mikulecky (2011), the overarching goal of an SLE is to enable learning in the right place with the right resources, presented at the right time. However, achieving these objectives requires more than technological advancement, it also requires the active involvement of users. To simultaneously provide users with an engaging experience during a participatory process and ensure the collection of valid data for the SLE concept, a web-based interactive survey tool named "DuoLero" was developed. DuoLero generates an Artificial Intelligence (AI)-produced image, which can be evaluated according to various factors. This way the tool collects users' individual preferences which then can be included into the SLE design.

DuoLero was created using a human-centered design process (DIN EN ISO 9241-210), including a literature review and the subsequent design and implementation phase. It can then be provided to participants on a device e.g., a tablet. There, a series of either/or questions are presented to capture the learning space preferences of the users (e.g., separate vs. open area). They thus go through several questions, each with two possible answers, from which exactly one must be selected. After the final question, DuoLero creates an AI image of the ideal SLE based on the given answers. The tool provides an interface that requires no prior knowledge of AI prompting. A predefined backend prompt ensures that users’ selections are consistently translated into visually and stylistically uniform images, allowing them to easily visualize their preferences without concern for technical details. This paper describes the development process and the design of DuoLero in more detail.

Besides collecting preferences, the authors also explore how users perceived the SLE visualization. Instead of using inflexible questionnaires, the evaluation was integrated directly into the tool, turning a static survey into an engaging interactive one. Users rated the AI image with one to five stars and evaluated aspects such as accuracy and aesthetic appeal and indicate whether they would learn or teach in the depicted SLE. The paper therefore also addresses the iterative development of the activating evaluation process, which is based on a scientifically sound foundation.

DuoLero was publicly deployed to collect a large audience’s preferences. At the same time, usability was determined using the System Usability Scale (SUS). The resulting high SUS Score (87/100) confirmed that DuoLero meets user expectations of usability. Thus, a tool was created that not only provides an engaging experience for users but also allows to collect valuable research data.

The collected data supports further SLE research by capturing preferences that can inform the design of future learning environments.The next step involves collecting extensive data from both teachers and learners in order to obtain a comprehensive understanding of learning space preferences. Also the visualization process can be seen as an initial step of many possibilities. Next, a 3D representation of the SLE could be created to function e.g., as an architectural prototype. At the same time, the modularity of DuoLero offers the potential to add further evaluation criteria, exchange questions, or conduct surveys in any other research context.

Keywords:

Smart Learning Environment, Research in Education, Human-Centered Design, Web-Based Survey Tool.