With the recent growth in interest in computing education for younger children, concrete representations have enabled children to understand concepts that, in abstract form, may otherwise be considered too advanced. Introducing children to computing concepts through concrete representations may be necessary at younger ages, but there is also a need for effective methods for progressing from this towards abstract understanding. Concreteness Fading has previously been applied in computing contexts with some success, but questions remain about how to design learning environments that most effectively support this progression, including the role of intermediate representations. This paper presents an augmented reality (AR) environment that uses the Concreteness Fading approach to teach introductory topics in computer networks. An in-school experiment with 59 children aged 9-10 investigated the effect of three distinct methods of fading representations from concrete to abstract, using a between-groups pre-test / post-test design. Three groups were compared: concrete > concrete > abstract (“two-step”), concrete > intermediate > abstract (“three-step”), and concrete > concrete/intermediate > intermediate > intermediate/abstract > abstract (“five-step”). The results showed a statistically significant difference between the two- and three-step groups in favour of the three-step, confirming the value of explicitly linking mutual referents via an intermediate representation when moving from concrete to abstract. No statistically significant difference was found between the three- and five-step groups, suggesting that increasing the number of intermediate representations does not significantly improve the learning outcomes. We discuss the implications of these findings for the design of educational technology for children.

Computing concepts are often introduced to children using interactive concrete representations, such as Computer Science Unplugged (CSU) activities - an approach that may positively affect attitudes towards computing when applied in extra-curricular settings. However, within the classroom, children may be required to interact with abstract representations, e.g. during programming tasks, and little is known about how to design CSU activities that support this. Prior work in mathematics and computing suggests that the Theoretical Model of Concreteness Fading (TMCF) may effectively scaffold the progression from concrete to abstract and improve learning outcomes. This paper aims to investigate one of the TMCF recommendations: the use of physical manipulatives, motivated by the growing body of evidence that suggests that virtual manipulatives can, in many contexts, be equally effective while providing several design advantages. An experiment was conducted with 48 children to compare two alternative implementations of Concreteness Fading learning materials: physical concrete (Playmobil toys) vs. virtual concrete (an AR iPad app), on learning gains and attitudinal change. Using these materials, children were taught the fundamentals of network structure and internet routing. Results indicated no significant difference between groups on either measure.

HCI researchers are increasingly interested in the evaluation of educational technologies in context, yet acknowledge that challenges remain regarding the logistical, material and methodological constraints of this approach to research [ 18 , 53 ]. Through the analysis of the authors’ contributed thematic research vignettes, the following article exposes the practical realities of evaluating educational technologies in school settings. This includes insights into the planning stages of evaluation, the relationship between the researcher and the school environment, and the impact of the school context on the data collection process. We conclude by providing an orientation for the design of HCI educational technology research undertaken in school contexts, providing guidance such as considering the role of modular research design, clarifying goals and expectations with school partners, and reporting researcher positionality.

In primary education, concepts are commonly introduced through concrete instantiations, such as physical manipulatives and kinaesthetic activities, with an expectation that learners will gradually move towards working with abstract representations. There has been considerable research in subjects such as mathematics on how children can move from working with concrete to abstract materials, but relatively little research on how this can be achieved in computing, which has recently become a more prominent subject at primary level. This paper reports on the design and evaluation of a low-fidelity prototype learning environment that aims to teach children aged 9--10 about a key computing concept (internet routing), using a concreteness fading approach commonly applied in mathematics. An empirical study with 59 children showed that those following a concreteness fading progression scored significantly higher on a post-test than those using a concrete only prototype, and had an increase in positive attitude towards computing in line with alternative approaches. We highlight the potential for an augmented reality implementation of the prototype to support investigation of further key questions raised by this research.

How can educational technologists most effectively support children in their development of computational thinking? This project attempts to integrate recent evidence from embodied cognitive science with concreteness fading: an instructional theory popular in mathematics and science education. Through a series of classroom studies, I will investigate the effectiveness of the concrete-pictorial-abstract sequence, and the relative benefits of physical and virtual implementations of the concrete stage. Results will be used to guide the development of an interactive learning environment for use in primary school classrooms.