Purpura et al. (2017), proponents of dialogue reading as a pedagogical tool for early grade teachers and parents, argue for its broader use. The authors note that it is underused and could serve as a vehicle to expand vocabulary and, with that concept development in an intersect of the disciplines of science, technology, engineering and mathematics (STEM). To develop a dialogue reading programme (DRP), which aims to advance STEM literacy in the early grades, requires an assemblage of: (1) theoretical and scientific knowledge of how young children learn (e.g. Clements & Sarama 2016; Dehaene 2020); (2) repetitive input from experts in the applied fields of STEM and education, as well as parents and children (e.g. John et al. 2018); and (3) careful investigation of existing examples of teaching material (particularly stories) in the STEM disciplines (Clements & Sarama 2016). Although a few theoretical frameworks for teaching STEM literacy exist (Chu, Martin & Park 2019; Greca Dufranc et al. 2020; Kelley & Knowles 2016), the resources with specific instructional guidelines for early grade teachers and parents are limited (Chu et al. 2019). Also, these existing frameworks mostly focus on elementary grade STEM education and explain how teachers can plan to integrate the disciplines of STEM in lessons and not specifically on a framework to ground the development of teaching material that can support teachers in doing so. There is, furthermore, no existing conceptual framework that grounds the development of a DRP to support STEM literacy education in the early grades. This is why Clements and Sarama (2015) highlights the idea that research informed programmes that can scaffold young children’s learning, should take precedence in early STEM literacy education.

The purpose of this article is to describe the conceptual framework that was used to develop a DRP for early STEM literacy. I set out to respond to the question: How can the design of a conceptual framework serve as theoretical groundwork for the development of a DRP for early grade STEM literacy education? I propose that an interdisciplinary conceptual framework can serve as the bedrock from which how early grade South African children can attain the STEM skills included in the new Digital Skills for All Curriculum (DSfAC) through ‘dialogue reading’ can be explored. Purpura et al. (2017) propose that:

The reading activity becomes a dialogue, rather than a listening activity. In a developmentally appropriate DRP, children can ‘participate’ in an interactive fashion and take part in the discussions between the characters in the books. Figure 1 shows the characters that feature in the three picture books.

The South African Department of Basic Education (DBE) aims to strengthen STEM literacy in the national curriculum through a DSfAC for Grade R – 9 (DBE, in press). The DSfAC includes computational thinking (CT), coding and robotics in the term ‘digital skills’. Bybee (2010:996) proposes that the purpose of STEM education is to equip children with the necessary skills to ultimately establish ‘a STEM-literate citizenry prepared to address the grand challenges of the 21st century’. Unfortunately, ‘studies show that the country lags behind in the information technology skills needed for the digital revolution’ (https://www.gov.za/ss/speeches/president-cyril-ramaphosa-2021-basic-education-sector-lekgotla-25-feb-2021).

The purpose of the DSfAC (DBE, in press) is to respond to a proposition, such as Bybee’s (2010), equipping children with skills required in a 21st century society and to build economic communities that can flourish in a fourth industrial revolution (4IR) world. Increasingly, STEM skills, including ‘digital skills’, are foundational in a technology-saturated society. Yet, less than 9% of South African learners are currently enrolled for school subjects that develop their ‘digital skills’ at Further Education Training (FET)-level and no digital skills are currently included in the early grades curriculum (DBE, in press).

The DSfAC consists of four pillars: (1) CT skills and coding; (2) data and information management skills; (3) internet and e-communication skills; and (4) application skills, with CT as the basis from which the programme will be approached (DoE, in press:8). In the foundation phase (Grade R – 3), the four pillars will be integrated into the existing subjects of literacy, numeracy and life skills.

The first pillar, which focuses on CT and which also forms the foundation of the curriculum, is situated within the STEM disciplines because CT can be readily applied in problem solving in STEM. Lu and Fletcher (2009) suggests that CT should be used to prepare young children for coding and other digital technology applications such as robotics. I would add that there should be coherence within the STEM learning areas and that an organising pedagogical mechanism could, for example, be CT. Wing (2008) also suggests introducing CT as a formative skill together with literacy and numeracy which is what I aim for with the DRP for early grade STEM literacy.

The DSfAC defines CT as a problem-solving process that includes a number of characteristics, such as logically ordering and analysing data and creating solutions using a series of ordered steps (or algorithms), and dispositions, such as the ability to confidently deal with complexity and open-ended problems. CT requires the use of abstraction, decomposition, pattern recognition, data representation, generalisation and modelling (DBE, in press:6).

The envisaged curriculum concentrates on teaching children how to solve real world problems and to develop children’s vocabulary, allowing them to describe possible solutions to problems and to reflect on their own solutions as well as those of others. That is why storytelling and interactive discussions about relevant STEM concepts are vital. In the foundation phase, children will also be taught to follow the scientific/engineering design process to help them plan and build simple robots, associated with movement, by following a verbal (natural language) or pictorial algorithm/plan such as the programming language tool, ScratchJr (Bers 2018; https://www.scratchjr.org/; https://scratch.mit/edu/boost).

Yet, apart from this, no teaching materials are recommended in the curriculum with which to apply the theoretical ideas described in the curriculum. The purpose of developing a DRP for early STEM literacy is to address this void by providing early grade teachers and parents with a practical tool to scaffold CT, vocabulary development and concept learning before they are exposed to the robotics curriculum at school. I argue that a practical teaching tool like the DRP can be utilised to initiate young children into the world of CT and specifically to achieve the foundational goals of the DSfAC. The DRP is designed to serve as examples of how children can create, invent, experiment, reflect and discuss solutions to real life problems and to develop their creativity and their critical thinking that the DSfAC requires.

Dufranc et al. (2020:5) propose that a framework that underpins didactical/instructional material should ‘address pedagogy and content within an integrated STEM education’. To develop a conceptual framework that could form the bedrock for a DRP for STEM literacy in the early grades, I focused on integrating (1) theories about how learning through language, and theories about CT, coding and other digital skills; (2) the advice of practicing experts in both education and the applied fields of engineering, coding and programming; and (3) examples of existing teaching materials.

I implemented an iterative participatory approach to the research plan for this article (Holt & Asagbra 2021; John et al. 2018). At the outset of the ongoing research project that was started in 2020 and in which I aim to write a series of three children’s books that can be utilised in the early grades to teach STEM concepts, I discussed many ideas with parents of young children and early grades teachers. During these conversations, most of the discussants referred to the components of digital skills which are modelled in existing stories on television. An existing DRP for early grade numeracy (Purpura et al. 2017, in press) also came up in the discussions regarding the layout and style of the books. I aimed to integrate the components of these discussions with existing theory about how children learn. I am familiar with such theories in the fields of developmental cognitive psychology (e.g. Carey 2009; Carey, Zaitchik & Bascandziev 2015; Clements & Sarama 2016) and developmental cognitive neuroscience (e.g. Ansari 2017; Dehaene 2011, 2020), specifically theories that describe the role of language in learning (e.g. Chomsky 2014; Dehaene 2009, 2013; Gallistel & King 2011; Gopnik & Meltzoff 1997; Levine & Baillargeon 2016; Vygotsky 1978, 1986) and with the idea that a theoretical/conceptual model should form the bedrock of design for educational material (Bezuidenhout 2018; De Villiers 2015). Yet, I did not previously consider the importance of existing story examples, or the value of an iterative participatory approach (Holt & Asagbra 2021; John et al. 2018) that incorporates the knowledge of many role players such as parents, teachers and children. The initial discussions with parents and teachers led to the composition of a conceptual model to ground the project by amalgamating theoretical knowledge of how children learn with the ideas of parents, teachers, children and other community members – which included references to for example, artificial intelligent robots and their application of digital skills and other STEM components in existing stories.

Figure 2 shows how the intersect of the three core components – theory, participation of various role players and existing stories as examples – collectively outlined the process of designing five aspects that constitute the DRP. Each aspect is discussed in the next section:

The series was named Robot Dilemmas and consists of a collection of three dialogic reading picture books. It was designed to scaffold children’s CT, and their coding and programming vocabulary while they identify with story characters; the characters solve problems, create new things in an experimental fashion and think logically. Figure 3 shows an example scenario in which learners can identify with the characters who must solve a problem. In this picture, Alex is moving too fast on a hiking trip with his two friends. As a result of the way in which Ben programmed the little robot, Alex keeps tripping over rocks and tree roots. Subsequently, Kate explains to Ben how the code should be adjusted so that the problem can be solved.

The books can serve as pathways to get acquainted with new words and their meaning, with which to build concepts and develop an intuition for coding language. Through repeated dialogic readings and discussions, children can internalise new vocabulary, preparing them for the digital (coding and robotics) curriculum which will soon be introduced into foundation phase classrooms. In the following section of this article, I describe the three central components of the conceptual framework and I report on how each component of the conceptual framework featured in the development of the Robot Dilemmas book series.

The theories included in the conceptual framework are an amalgam of various perspectives on (1) how children learn through language (including the use of dialogue reading); (2) the principles of CT, coding and other digital skills; and (3) how the language skills and STEM literacy were integrated in the stories.

Over the last few decades, a shift has taken place in how educational research is related to practice. Previously, it was thought that research yielded insights into educational practice and that findings could be applied in practice through a process of translation from research results and daily practice (Snow 2015). The assumption was that if the research process and methods were sound, applying the findings simply required interpretation and translation for application. Recently, though:

Both the researcher and practitioner’s knowledge is now judged to be of equal importance to improve educational outcomes. In the instance of this project, practitioners include a variety of role players, such as teachers, parents, illustrators and linguists.

The second core component of the conceptual model is the utilisation of an iterative participatory approach (Holt & Asagbra 2021; John et al. 2018; Purpura et al. 2021, in press) to involve various practicing role players in the development of the DRP. The goal of the participatory approach is to capitalise on the experience and expertise of each participant, both in the field of research and practice. By including the various contributors of the interdisciplinary team, consisting of education and developmental psychologists, teachers, children, parents and specialists in the fields of engineering, coding and programming, I designed the three dialogic reading books for young children.

Parent pairs who have children in preschool and of which at least one parent is a specialist in programming and engineering, were consulted on the story lines, characteristics of Ben, Kate and Alex, vocabulary use and how constructs are portrayed in the stories. Parents and teachers provided names of example stories (either on television or reading books) which were used to choose the story topics, design the characters and to inform the illustrators. Several conversations with linguists, researchers, teachers and parents took place in the preparation of the work, to ensure a discourse that is linguistically correct, yet appropriate for young children. For the illustrations, a local illustrator for children’s books was identified to contribute to the design of the characters and the development of the illustrations to support the texts.

The third core component of the conceptual model is the systemic utilisation of existing examples of teaching materials. Through the iterative participatory approach, parents and teachers indicated that Rusty Rivets, Puppy Dogs, Paw Patrol, Cars2 and Blaze are stories on television where little robots are presented with problems. Some of these stories contain a say-along-phrase to teach basic vocabulary and ideas, which was a concept that I incorporated in the DRP. Other stories have characters with distinct features that contribute to children’s understanding of required concepts. I used this notion by incorporating the already mentioned ideas of ‘input’, a ‘toolkit’ and a ‘lens’. The story lines are also age-appropriate and clearly portray the design process, similar to what happens in existing stories. Recent children’s books in this field of education (Buarque, Roberts & Marwala 2019; Giridhar 2019; Negishi 2018; Smibert 2018) were also used as examples.

Lastly, an existing DRP for early numeracy, The Little Elephant’s Big Adventures (https://earlymath.ericson.edu/families-are-pivotal-in-designing-math-picture-books-for-children/), designed by Purpura and his team at the Purdue University (Purpura et al. in press), also contributed significantly to the structure and layout. These are dialogic reading books to develop mathematics-specific language and they consist of text and three example questions on each page which the reader can use to ask questions and prompt responses. The same structure was used in the design of Robot Dilemmas.

The combination of the three intersecting core components jointly contributed to the development of five key features of the DRP: (1) story topics; (2) characters; (3) vocabulary; (4) constructs; and (5) illustrations.

This article has argued that dialogue reading is an effective way to develop vocabulary and concepts consecutively. The DRP which I have discussed was designed to teach the constructs outlined in the DSfAC. This curriculum will soon be implemented in the early grades of South Africa. Computational thinking, coding and programming robots are digital skills required in the 21st century, and are also skills that are interwoven in the disciplines of STEM. Although theoretical frameworks exist that describe how teachers can plan and teach STEM literacy in a connected fashion, conceptual frameworks for the development of teaching material to support their teaching, such as a DRP have not yet received attention. To this end, I have proposed the interdisciplinary conceptual framework which served as bedrock for the development of a DRP and which addresses the concepts introduced in the new South African digital skills curriculum. I argue that if one cannot define the concepts needed to teach CT, coding and other STEM constructs through reading, one cannot study how a DRP will influence children’s understanding.

By developing a conceptual model that grounds the design of teaching and learning material, I also proposed that a theoretical perspective alone is not sufficient and that theory should be supplemented by the perspectives of other role players though an iterative participatory approach and also by existing examples of stories used in the early grades. In the next phases of the current ongoing research project, the validity of the DRP will be assessed in a pilot study before conducting an experimental study to assess the workability and implementation of the DRP.