Towards tailored cognitive support in augmented reality assembly work instructions

Background Augmented reality (AR) is receiving increasing interest as a tool to create an interactive and motivating learning environment. Yet, it is unclear how instructional support affects performance in AR. Objectives This study sought to explore how varying the instructional support in AR can affect performance‐related behaviours of students with low cognitive abilities during assembly work. Methods A total of 90 Belgian secondary school students repeatedly executed four different realistic assembly tasks. Three levels of instructional support (low, medium, and high) in AR as well as a control condition with paper instructions with a high level of detail were systematically varied across tasks and participants. Results and Conclusions Multilevel regression analyses showed that AR instructions yielded lower assembly times and a lower perceived physical effort than paper instructions. Additionally, participants perceived tasks as less complex when given AR instructions with a high or a medium level of detail than when given a low level of detail. No effects of instructional support were established for other performance‐related behaviours, namely necessary assistance, error‐making, cognitive load, competence frustration, and stress. Effect sizes were small, at least among the instructional support conditions studied, yielding a limited base for adaptivity. Presumably, tailoring the instructional support in AR is only beneficial for highly complex tasks. The results might be useful for the design and implementation of AR in educational settings. Lay Description What is currently known about the subject matter Augmented reality is knowing increased use in education. Little is known about how to design effective instructional support in augmented reality. What this paper adds This study represents an initial insight into how to personalize instructional support. The impact of instructional support on augmented reality performance of students with low cognitive abilities was investigated. The study suggests that varying instructional support may lead to differences in performance outcomes. Implications for practitioners How instructional support is constructed may affect augmented reality learning. The results may inform the design of effective augmented reality learning environments for students with special needs. How individual, contextual, and task‐specific characteristics moderate the effectiveness of instructional support should be further investigate