Literature Review

Introduction

This chapter reviews relevant literature to summarise research that supports the aims of this study. It explores several themes to establish a comprehensive understanding of the field of computer game design and programming (CGD&P). First, I summarise the findings of reviews of research on CGD&P. Given its significance in the field and practical applications, I then outline key concerns raised by constructionist researchers in their efforts to support programming education, both broadly and within CGD&P. The review then addresses pedagogical approaches relevant to the thesis’s research questions. For the purposes of this chapter, I draw on a definition of pedagogy aligned with the socio-cultural approach of this study from Siraj-Blatchford and colleagues [-@siraj-blatchford_researching_2002] 1. I begin by examining structural pedagogies from the wider field of computer programming before focusing on those specifically addressing game making. Next, I review research on the potential and characteristics of social and cultural approaches to game making, covering topics such as pair and peer coding, coding clubs, competitions, and game jams. This section ends with an exploration of novel and robust pedagogies that enhance learner agency through initiatives such as the Fifth Dimension (5thD) interventions and the Connected Learning research programme. Finally, a significant focus is placed on the use of game design patterns (GDPs), highlighting their application in computing education. By addressing these themes, the literature review positions the current study within the broader academic discourse. Given the central rationale of the thesis, it specifically examines pedagogical approaches capable of fostering inclusive practices within game making communities for novice coders.

A summary of research on computer game design and programming (CGD&P)

In order to provide a broad perspective on overarching concerns in this area, this section first examines studies on computer game design and programming (CGD&P). Several reviews explore the motivations, processes, and benefits of game development for learning [@denner_does_2019; @earp_game_2015; @hayes_making_2008; @kafai_constructionist_2015-1; @gee_video_2016]. Kafai and Burke’s [-@kafai_constructionist_2015-1] review synthesises 55 relevant studies within the framework of constructionist gaming, focusing on different strands of CGD&P’s potential as well as the barriers to participation (explored in Chapter 1). The most developed section concerns how CGD&P research addresses the development of personal knowledge and skills. Key topics include computer science knowledge and programming, mathematical and scientific competencies, and technical abilities that enable participation in the information society. Additionally, social aspects such as pair programming, social skills, self-reflection, and cultural awareness are addressed. The authors note limitations in how many findings present details regarding the pedagogical approaches used; however, studies that do include such details are not included in the review [@illingworth_review_2017].

Hayes and Games [-@hayes_making_2008] also adopt a broad approach in their review, identifying four main motivations for CGD&P: learning computer programming skills, deepening knowledge of other curriculum subjects, involving more girls in computer programming, and using game design to understand design concepts. Other reviews focus more narrowly on specific themes. While Gee and Tran [-@gee_video_2016] discuss the variety of tools available for game design, Bermingham and colleagues frame CGD&P competencies as essential 21st Century Skills [@bermingham_approaches_2013], a focus echoed in a comprehensive review on digital making, which includes game development [@lamas_making_2019]. In their review of CGD&P’s potential to encourage collaboration, Earp and colleagues [-@earp_learner_2013] found that analysis of collaboration primarily focused on peer review and other forms of feedback.

A review of CGD&P by Denner and colleagues [@denner_does_2019] examines CGD&P’s effectiveness in promoting computing science learning and motivation, breaking this broader concept into subcategories of programming knowledge, problem specification, and design. Their findings remain cautiously optimistic but highlight limitations in existing research, including unclear motivations, insufficient detail on the research process, and limited demographic information. While Denner et al. [@denner_does_2019] outline three emerging strands of pedagogical interest—design-build-test, step-based instruction, and social pedagogical approaches—due to limitations in source research, these strands are described only broadly. Given these limitations and others noted in previous reviews, this chapter will conduct a review of CDG&P research that provides more detailed pedagogical insights. This review will also incorporate perspectives from broader computing and digital making pedagogies. Before undertaking this more detailed review, it is important to first acknowledge the breadth of innovations and research contributions made by the constructionist school. This overview serves to highlight the significance of this body of work and to outline key themes that will be explored in depth.

An overview of constructionist approaches

The work of constructionist researchers forms a crucial foundation for the research landscape in both CGD&P and broader programming education. As such, many of the themes relevant to this literature review originate in constructionist research, often emerging from the activity of the MIT Media Lab [@semenov_seymour_2017]. Examples include the pioneering work of Papert on LOGO 2, his innovations in programmable LEGO drawing turtles [@papert_mindstorms_1980; @papert_turtles_2002; @papert_childrens_1993], Kafai’s early investigations into game programming by children [@kafai_minds_1994], Resnick’s contributions to creative computing, including physical computing [@resnick_digital_1998], and Scratch—a more generalised multimedia authoring tool [@resnick_reflections_2005]. Kafai and Burke’s [-@kafai_constructionist_2015] framing of CGD&P as constructionist gaming underscores the dominance of this perspective within the field. Consequently, many researchers working in this area align their studies with constructionist principles [@harel_caperton_toward_2010; @repenning_scalable_2010; @weintrop_computational_2016; @kynigos_children_2018].

Despite this influence, the ethos and theoretical underpinning of constructionism remain difficult to pinpoint [@laurillard2020significance, p.29]. This lack of clarity may partly result from the varied interpretations of constructionism as an “epistemological paradigm, a learning theory, and a design framework” [@kynigos_constructionism:_2015]. The following sections position the concerns of this thesis within the existing body of constructionist research before expanding the focus to explore its principal strands.

Constructionist pedagogies: Microworlds, design principles, and resulting toolsets

Constructionist Pedagogies: Microworlds, Design Principles, and Resulting Toolsets

Early constructionist research by Papert and Turkle [-@papert_epistemological_1990] examines diversity in coding approaches as a means to counter potentially alienating abstract methods. The authors describe an alternative, more tangible computing pedagogy they call bricolage, a craft-oriented approach in which participants become intimately acquainted with their tools and materials. Bricolage approaches feature an iterative process, close engagement with the code, strong links between function and form, and a persistent focus on maintaining a concrete understanding of outcomes, even if this comes at the expense of programming efficiency or code neatness. This approach aligns particularly well with tangible physical and digital objects, especially those designed for sharing within communities [@kafai_constructionism_1996-1].

Laurillard [-@laurillard2020significance] describes constructionism as a distinctive pedagogy rooted in Papert’s vision of carefully scaffolded coding environments (Microworlds) that support publicly shareable projects. Microworlds are simplified computational models where students experiment with abstract concepts (e.g., mathematics or physics) in a concrete manner [@rieber_microworlds_2004; @harel_constructionism_1991; @papert_mindstorms_1980]. Specialised, task-specific programming languages reduce the complexity of code compared to fully featured programming languages [@kong_providing_2022]. The use of Microworlds within traditional school settings presents challenges, as there is a risk that their potential may be diminished to mere instructional tools targeting teacher-selected curricular concepts [@hoyles_microworldsschoolworlds_1993]. Papert [-@papert_turtles_2002, p.17] also warns against institutional dilution, which could undermine their ability to remain “exploratory, playful, personally meaningful.”

Also working within the MIT Media Lab, Resnick and Rosenbaum [-@resnick_designing_2013] outline key constructionist design principles that support a process of designing for tinkerability. Fundamental characteristics include providing immediate feedback, fluid experimentation, and open exploration. The principle of simplifying complex and obscure processes in digital production tools to enable learners to quickly engage in creative processes is described metaphorically as low floors. Other principles include high ceilings, reflecting the importance of ensuring that projects are not constrained in terms of complexity if learners wish to pursue more ambitious work, and wide walls, which highlight the value of enabling a diversity of media genres, project types, and approaches [@resnick_reflections_2005; @resnick_scratch_2009].

A concern particularly relevant to this study is the tension arising from structural support or scaffolding intended to uphold the principles of wide walls [@bruner1974communication]. For example, the constructionist concept Constructopedia was developed to provide choice-based support for diverse project pathways [@papert_technological_1995]. Constructopedias function as small-scale, online encyclopedias of design elements and resources, facilitating concrete implementation while serving both as practical tools and sources of inspiration. Nichols [-@nichols2007idea] outlines challenges associated with attempts to create resource repositories based on Constructopedia principles, including contextual limitations within certain repositories [@carbonaro2004using], the inability of participants to contribute their own resources, leading to a diminished sense of ownership, and the overly broad scope of some repositories [@nichols2007idea].

The principles of the Constructopedia concept are partially integrated into Scratch, the MIT Media Lab’s multimedia authoring tool for novice coders. The wide walls principle is reflected in the availability of a vast asset library within the Scratch authoring tool. Additionally, all products created within the online community can be remixed by other users. However, extracting specific features from existing projects through remixing is not straightforward, as the embedded nature of these features complicates the process [@amanullah_evaluating_2019]. The two approaches, remixing and Constructopedias, highlight a tension between providing inspiration and ensuring learners have sufficient technical details to implement relevant features in their own projects.

Coding clubhouses and cultural programmes

Constructionist research is often conducted in community settings [@resnick_computer_1996], with a notable example being the first Computer Clubhouse, an after-school club based in Boston’s computer museum [@resnick_computer_1998; @peppler_computer_2009]. Seymour Papert, the founder of constructionism, [-@papert_mindstorms_1980, p.149] was influenced by his observational research on community organising within Samba schools as a form of mutual learning environment, highlighting the value of settings that are “real, socially cohesive, and where experts and novices are all learning.”

Bruckman [@bruckman_community_1998, p.51-52] emphasises the importance of a constructionist culture in supporting experimental processes enabled by technological tools. Building on this concept, Bruckman and Zagal [-@zagal_samba_2005] later formally compared the cultural learning components of a Samba school to activities within a computer clubhouse. To conduct this analysis, the researchers applied socio-cultural research on community practices [@lave_situating_1991] to examine social practices such as the importance of showcase events for sharing created work, the flexibility in participation methods, and the value of diversity in participants’ skill levels and backgrounds.

MIT researcher Roque [-@roque_family_2016] integrated family members directly into the making process to address barriers to computer coding as part of the Family Creative Learning (FCL) program. FCL incorporates both constructionist and socio-cultural ideas within face-to-face sessions, a creative making programme using Scratch, and playful physical processes and materials. Roque builds on the work of Barron and colleagues [-@barron_parents_2009] on parental roles in a digital making environment to inform facilitators in helping parents and children develop their participation within community activities [@roque_becoming_2018]. Barron et al. [-@barron_parents_2009] identified social and cultural behaviours of parents in settings involving informal technology use, categorising these behaviours into specific roles, including teacher, project collaborator, learning broker, non-technical consultant, and learner. Roque effectively demonstrates the importance of exploring these collaborative roles among parents and facilitators, as well as the value of designing social, collaborative, and reflective activities to complement the more technological aspects of making activities. The research team developed a detailed guide for replicating the programme [@leggett_family_2017] 3.

Constructionist framings of computational thinking and dimensions of fluency and agency

Constructionist research has contributed significantly to understanding computational thinking (CT) and the role of fluency and agency within digital making. Regarding computational thinking, Tedre and Denning [-@tedre_long_2016] highlight Papert’s foundational applications of CT and caution against more recent definitions that prioritise formal abstractions [@wing_computational_2008]. Differentiating between concrete and abstract approaches is particularly valuable for examining how these contrasting perspectives influence computing education. Wing’s [-@wing2011research] influential conceptualisation of CT emphasises various abstraction processes and has been widely incorporated into learning resources targeting educators [@bbc_bitesize_introduction_nodate; @dong_prada_2019]. Her framework identifies four key pillars of CT: decomposition, pattern recognition, abstraction, and algorithmic thinking. This perspective suggests that understanding the core principles independently of the coding context enhances their applicability beyond computing. Many computer science educators have leveraged this broader applicability rationale to support the mainstream integration of computer science into education [@guzdial_paving_2008].

Constructionist researchers Brennan and Resnick [-@brennan_new_2012] responded to the limitations of Wing’s definition of computational thinking (see [@cuny_demystifying_2010]) by adopting a grounded, situated approach to mapping the potential learning dimensions of students designing and coding collaborative, creative computing projects. Their resulting framework for computational thinking identifies three core categories: computational concepts, computational practices, and computational perspectives. Examples of these elements include concrete code concepts (e.g., loops, conditionals, and sequences) and practices (e.g., debugging, iterative practice, reusing, and remixing).

The broader role of community is reflected in the third category of computational perspectives, which includes expressing, which refers to creating projects that enable self-expression within a peer community; connecting, which involves engaging with others through shared computational activities; and questioning, which encourages a critical approach to technology. Lye’s extensive review of teaching computational thinking [@lye_review_2014] incorporates Resnick and Brennan’s definition as its foundation, indicating the widespread use of this applied, context-driven perspective. By reconnecting with Papert’s vision of computing projects embedded within community settings, this work represents an important reaffirmation of a grounded, flexible perspective on computational thinking [@tedre_long_2016; @denning_remaining_2017].

The applied framework of Resnick and Brennan builds upon previous work by Papert and Resnick [-@papert_technological_1995] on technology fluency. Fluency as an attribute appears across multiple strands of constructionist research, described in different contexts as technical fluency [@papert_technological_1995], digital fluency [@resnick_scratch_2009], gaming fluency [@peppler_gaming_2009-1; @kafai201221], and computational fluency [@resnick_computational_2018]. Similarly, the work of Kafai and colleagues examines collective agency [@kafai_learning_2008] and later collaborative agency [@kafai_collaborative_2011; @kafai2012collaborative], focusing on collective work in after-school computer clubs and the online Scratch community. The researchers apply theoretical concepts such as communities of practice [@wenger_communities_1998] and agency formation during collaborative knowledge production [@scardamalia_higher_1991-1] to explore collaborative agency. They observe learners distributing responsibilities as they develop collaborative projects. However, their discussion of the process remains brief, focusing primarily on general online collaboration within the Scratch community rather than face-to-face settings. Unfortunately, Kafai’s later work on game making does not significantly expand upon this promising strand of exploring agency.

Gaps remain in constructionist research regarding pedagogical approaches. Despite Kafai’s emphasis on the importance of a situated and critical approach to coding practices [@kafai_revaluation_2022; @kafai_theory_2020], some critiques highlight a lack of explicit pedagogical structures. Vossoughi [-@vossoughi_making_2016] critiques constructionism from a socio-cultural and egalitarian perspective, pointing to the absence of intentional forms of pedagogy. Similarly, while constructionist approaches acknowledge the importance of self-expression within a peer community as an aspect of fluency (see above), they frequently lack a conceptual underpinning.

Vossoughi [-@vossoughi_making_2016] attributes this gap partially to the increasing reliance on researcher-created toolkits and communities that shape personal understandings of knowledge [@vossoughi_making_2016]. When pedagogy is incorporated into recent constructionist studies, it often appears in the form of broad principles of design thinking or general project-based learning strategies [@resnick_scratched_2012; @resnick_lifelong_2017]. A later section examines these pedagogies in greater depth.

It seems likely that this deficit arises more from omission than intentional design, given the similar critiques made by constructionist researchers themselves, who caution against overly technical approaches and toolsets at the expense of expressive potential within community settings [@resnick_seeds_2020; @resnick_coding_2020]. Kafai and Burke have also advocated for greater research into the social and cultural dimensions of game making. These gaps in CGD&P research, which has been significantly shaped by constructionism, may reflect limitations in both analytical processes and theoretical concepts employed by constructionist researchers.

Aligning with the primary research question, which investigates how CGD&P research can be enriched through socio-cultural approaches, the next section examines varied pedagogies that offer contributions to this line of inquiry.

Pedagogies to support game making via coding

One of the main themes of this review is an exploration of pedagogies that support CGD&P. The following sections focus on pedagogies relevant first to general programming, and then more specifically to game making.

Explicit teaching: Step-by-step instruction and principles-first teaching of computational thinking

In Denner et al.’s [@denner_does_2019] review of CGD&P, stepwise learning approaches were observed in 40 out of 68 studies 4. Within step-by-step instruction or tutorials, educators typically guide learners in using tools to achieve pre-set goals, embedding underlying principles of computational thinking and concepts within this instruction so that learners absorb them through active engagement. An alternative instructional strategy is the principles-first approach [@repenning_scalable_2015], informed by Wing’s [@wing_computational_2008] advocacy for explicit teaching of decomposition, pattern recognition, abstraction, and algorithmic thinking. Given the abstract nature of computational thinking (CT) principles, this approach applies to varied domains of coding.

Wing’s perspective on computational thinking is more theoretical than applied and invites discussion on effective methods for delivering principles-first approaches. Grover [@grover_computational_2017; @grover_computational_2013] and Guzdial [@guzdial_learner-centered_2015] outline explicit techniques that leverage computer programming as a mechanism for teaching CT principles. Bell and colleagues [@bell2019constructing] investigate unplugged activities, which introduce abstract CT concepts without using computers or coding. Other research examines challenges in teaching computational thinking principles within non-computing disciplines [@dong_prada_2019].

The remit of this literature review does not extend to a full critique of the validity of Wing’s strand of computational thinking within computer science or other disciplines 5. However, while a principles-first approach, which emphasises understanding foundational concepts before practice, may appear logical, it can exacerbate foundational barriers to participation when learners encounter complex material, as discussed above [@papert_epistemological_1990] and in Chapter 1. Consequently, the following sections prioritise pedagogical practices that lead with or integrate concrete exploration of coding processes.

Design frameworks using stages & project based learning (PBL)

As identified by the systemic review of CGD&P [@denner_does_2019], the design-built-test pedagogy is common in this field found in 30 out of 68 studies. The design-build-test exemplar cited in that review highlighted the approach of Globaloria programme. Globaloria structures project work around an iterative cycle of design stages namely: Play, Plan, Prototype, Program, and Publish. Many similar frameworks exist in diverse areas of production including computer science [@pereira_design_2018], engineering [@winarno_steps_2020-1], design processes [@dam_5_2024] and project based learning. Addressing the domain of design thinking, Figure 2.x includes a graphical representation of design thinking stages from the Institute of Design at Stanford [@dam_5_2024]. The stages are be described in the following way: Empathise - learn about the audience; Define - sharpen key questions; Ideate - Brainstorm and create solutions; Prototype - build representations of one or more ideas; Test - test ideas and gain user feedback.

Figure 2.x. Design thinking via design stages model from Stanford dschool (left) & Creative Learning Cycle by Mitchell Resnick{width=100%}

Similarly, to illustrate the iterative nature and importance of community in of a design approaches to making Resnick advocates the use of a creative cycle model for educators [@resnick_lifelong_2017]. The five circular stages are; Imagine, Create, Play, Share, Reflect and returning to Imagine once more. Within the context of computing education Resnick [-@resnick_scratched_2012] describes the foundations of the design-based approaches in education as; engaging in design activities, exploring personally meaningful topics, collaborating with others, and deepening understanding through reflection. While research on the use of design thinking in education is broadly supportive, finding promise to develop varied forms of creative planning and collaborative skills [@luka_design_2019; @lor2017design], there are limitations within the field of design thinking and education that stem from lack of clarity regarding terms, information on how to implement design stages, and vague theoretical justification of the overall process [@micheli_doing_2019]. To address these limitations we can turn now to the fields of project based learning which shares the stage based characteristics of design thinking but which has been developed and analysed in line with socially constructed theories of learning 6.

The field of project-based learning (PBL) encompasses a wide range of research in educational contexts, exploring pedagogical approaches aligned with the social learning focus of this thesis. PBL approaches also incorporate an iterative design model [@jia2023design], and existing research provides detailed analyses of the processes and rationale involved at each stage. Broadly defined, PBL is an educational strategy that advocates: learner autonomy in project selection, which enhances motivation [@darling-hammond_powerful_2008]; authentic and shareable project outcomes, along with learning environments that encourage peer feedback and reflection [@gibbes_project-based_2014; @hung_activity_2000]; iterative project work, which supports student mastery; and challenging goals with structured <!– To align with the focus of this study I will focus on the following dimensions of PBL:

  • driving questions / challenge - involving chi
  • facilitation of group creativity
  • authenticity of processes –>

Significant challenges exist in undertaking PBL, including the practical constraints of timetabling and time pressures within formal settings [@marx_enacting_1997], restricted access to authentic resources [@thomas_review_2000], and variations in expertise and confidence in facilitation [@ertmer_essentials_2015-1]. One of the core difficulties for educators arises from the shift in perspective required to move from a traditional teacher-led approach, in which subject knowledge is directly instructed, to a facilitator role, where students take greater ownership of directing activities. Due, perhaps, in part to the use of learning environments and activities that prioritise the active construction of knowledge by learners rather than instruction-heavy models [@kokotsaki_project-based_2016], PBL is frequently misinterpreted as an entirely unstructured, pure-discovery approach [@kirschner_why_2006]. However, while no single framework for scaffolding exists, contextually relevant support remains a vital component of PBL [@hmelo-silver_scaffolding_2007].

Ertmer and Simons [-@ertmer_scaffolding_2005] highlight the value of differentiating scaffolding strategies to support teachers in delivering PBL effectively. Saye and Brush [-@saye_scaffolding_2002] distinguish between hard and soft scaffolding, with hard scaffolding providing structured, static support for planning and content organisation, while soft scaffolding facilitates dynamic, relational strategies such as responsive questioning and collaborative problem-solving. In a STEM learning context, Pitot and colleagues [-@pitot_establishing_2024] analyse a widely recognised project design rubric developed by PBL Works 7, which emphasises the development of complex project challenges, sustained inquiry, authentic engagement with tools, student agency, critique and revision, reflective learning, and publicly shareable outputs. Ertmer and Glazewski [@ertmer_essentials_2015-1, p.97] present a similar categorisation, highlighting the importance of group work facilitation, structured opportunities for reflection, and discipline-based argumentation, where project work helps learners explore subject-specific knowledge frameworks.

The authors also identify a tension in PBL design, balancing the benefits of student autonomy in sustaining engagement against the practical limitations of restricting participant choice, to ensure facilitators can provide adequate support without being overwhelmed. Leat [-@leat_enquiry_2017-1], in his analysis of school-based PBL through a sociocultural lens, emphasises the importance of integrating what Moll [-@moll_funds_1992] describes as funds of knowledge, resources that learners draw upon in both formal and informal settings 8. However, implementing such resources presents challenges for teachers, who may lack familiarity with the communities their students participate in [@gonzalez_funds_2007]. Leat [-@leat_enquiry_2017-1] suggests that activity theory provides a valuable sociocultural framework for examining these tensions within ecological learning environments.

Project-based learning frequently presents students with challenges framed as wicked problems, which require ongoing reassessment and adaptation to identify contextually appropriate solutions [@kleczek_wicked_2020]. Similarly, the challenge of designing effective scaffolding for PBL also meets the criteria of a wicked problem, requiring educators to adopt flexible, evolving strategies to accommodate emerging technologies and shifting learning contexts. Although the broader principles of project-based learning contribute valuable insights, literature regarding scaffolding specific to CGD&P remains limited, constrained not only by acknowledged gaps in sociocultural research [@kafai_constructionist_2015], but also by the inherently dynamic nature of the field.

Some studies explore how general PBL structures and iterative design stages can be applied to game-making, demonstrating their alignment with project-based learning approaches and their role in providing structural support [@kafai_constructionist_2015; @denner_does_2019]. For example, the goal of creating a game provides a motivating challenge that can be shared publicly, either in person or online, enhancing authenticity and engagement [@pitot_establishing_2024]. Simmons et al. [-@simmons_using_2012] briefly describe a five-stage design process to help students structure their project planning.

Similarly, researchers studying the Globaloria programme [@reynolds_formal_2013-1] found that the intrinsic motivation associated with game-making encouraged sustained engagement, increasing the likelihood that participants would refine, test, and revise their creations. However, many studies cited by Denner et al. as illustrating the design-build-test approach lack detailed exploration of students’ design processes, or fail to communicate scaffolding strategies for teachers and facilitators [@wang_relationships_2011; @wang_learning_2010-1; @robertson_influence_2013; @ke_implementation_2014]. While publication constraints may account for some omissions, this may also reflect genuine gaps in the research. For instance, the online supporting materials for the Adventure Author research by Robertson primarily focus on software usability rather than structured guidance on the design process 9. While general PBL structures contribute foundational support, more specific scaffolding approaches within CGD&P remain necessary and will be examined in detail in the following sections.

Use-Modify-Create (UMC) and half-baked games

The Use-Modify-Create (UMC) approach proposed by Lee and colleagues [-@lee_computational_2011] offers a promising strategy to mitigate user anxiety and demotivation associated with the complexity of coding games. UMC originates from research on game-making and robotics as mechanisms for enhancing computational thinking [@denner_computer_2012; @denner_using_2014; @werner_pair_2013; @werner_children_2014]. The model advocates remixing existing games as a scaffold for developing novice coders’ competence. Learners are guided to make progressively complex modifications, thereby becoming increasingly proficient in recognising and applying computational concepts and structures [-@lee_computational_2011].

In the Use stage, coders familiarise themselves with coding interfaces, code structures, and syntax. In the Modify stage, learners engage with pre-existing projects, deepening their knowledge of coding concepts and practices by adjusting the code to align with their own aims. In the Create stage, after developing familiarity with code design patterns introduced during the Modify phase, they advance to reproducing these patterns in independently created projects.

UMC enhances engagement by lowering technical barriers to participation and strengthening learners’ sense of ownership over their projects by allowing greater choice in defining final outcomes. A study involving five hundred 9- to 14-year-olds found that the UMC approach balances the need for structured learning while supporting student-led exploration [@franklin_analysis_2020]. Researchers also observed that students particularly valued UMC for its flexibility, choice, and agency in the process. Additional research comparing UMC with a starting-from-scratch approach found higher levels of student engagement in the UMC group [@lytle_use_2019], as learners spent more time actively manipulating code, incorporating personal touches that enhanced their sense of ownership over their projects.

These findings correspond with constructionist research, which highlights the motivational potential of allowing learners to design and publicly share code-based products [@kafai_constructionism_1996]. However, increased choice can also introduce two challenges: first, students may diverge from intended subject areas, and second, facilitators may experience stress in managing diverse and open-ended activities. Noss and Hoyles [@hoyles_pedagogy_1992] refer to the first issue as the play paradox, where increased freedom can lead to off-topic engagement. To address the second challenge, Lytle and colleagues propose replacing the open-ended ‘Create’ phase with structured extension options, shifting the emphasis from ‘Create’ to ‘Choose’ [@lytle_use_2019-1].

Half-baked games, proposed by Kynigos and colleagues [-@kynigos_half-baked_2007; -@kynigos_children_2018], are intentionally unfinished or contain deliberate deficiencies designed to motivate learners to modify the design or code and improve the game. The concept builds upon Papert’s work on Microworlds [-@harel_constructionism_1991; -@papert_mindstorms_1980]. Like Microworlds, half-baked games encourage malleability in directions that interest the learner [@kynigos_children_2018], while ensuring the learning process remains educationally productive. To achieve this, the designer of the half-baked game makes complex decisions that emphasise specific affordances of the game, encouraging learners to explore fundamental computational concepts 10.

Kynigos and Yiannoutsou [-@kynigos_children_2018, p.2] conducted a study which provided “tools affording them (13-15 year olds) with the role of game hackers,” allowing them to alter existing game code. The researchers identified progression in coding complexity among learners throughout the intervention, beginning with pattern recognition associated with reading code, and advancing to more sophisticated practices, including abstraction, structural organisation, and sequencing their own algorithms. In earlier work, Kynigos [-@kynigos_half-baked_2007, p.336] described the potential for half-baked artefacts to foster learner dialogue around coding challenges, framing them as “a communicational tool to shape a common language within the community,” highlighting their alignment with social and cultural pedagogical approaches.

Levels of abstraction, semantic profiles & PRIMM

The work of Sentence and Waite, culminating in a recent report on teaching strategies within UK computing education [@waite_teaching_2021], identifies a range of relevant pedagogical approaches, including levels of abstraction, semantic profiles, and PRIMM. The concept of levels of abstraction (LOA) supports both teachers and students in developing a hierarchical understanding of coding processes [@statter_teaching_2016; @waite_abstraction_2016; @waite_abstraction_2018]. While abstraction is a widely interpreted concept within computer science [@hazzan_reducing_2002], in this context, it refers to a continuum between overarching concepts and concrete outcomes (see Table 2.1).

Level Focus Example
Conceptual Level Thinking about the problem without programming Task - what is needed
Design Level Structuring / designing a solution What it should do.
Code Level Writing the actual code How it is done
Execution Level Understanding how the computer processes the code on a low-level (e.g memory use), or in a k5 context the outputs What it does.

Table. 2.1 Breakdown or Table of levels of abstraction. [@waite_abstraction_2018]

Guided by research highlighting the potential efficacy of abstraction in supporting programming practices [@cutts2012abstraction; @statter_teaching_2016], Waite and colleagues examined the relevance of abstraction awareness for primary school-aged learners [@waite_abstraction_2016]. While the process initially lacked a structured pedagogy, the authors identified potential in the ability to move between levels as a form of self-regulation and, more specifically, in activity at the design level, which enabled novice coders to make realistic judgments regarding code implementation based on their time constraints and abilities [@waite_abstraction_2018].

Semantic profiles illustrate the use of more concrete (high semantic gravity) language and more abstract (high semantic density) concepts and patterns as they develop in classroom situations [@macnaught_jointly_2013]. Research conducted by Curzon and colleagues [-@curzon_using_2020] within computing education examines the benefits of wave-shaped semantic profiles rather than flatlines, which either remain overly rooted in concrete examples or excessively abstract (See Figure 2.x for an example). This research highlights the importance of unpacking, exploring, and repacking ideas throughout a lesson, allowing a student’s understanding of a concept to gradually deepen as it is applied in practice and reconnected with abstract principles.

Figure 2.x - A semantic profile in wave form from [] {width=75%}

Later work by Waite and Sentance [@sentance_primm_2017] builds upon the ethos behind UMC, LOA, and semantic profiles [@macnaught_jointly_2013] to develop an applied pedagogy—PRIMM—for computing education. The motivation behind PRIMM is to address the tension between exploration-based learning approaches and those advocating greater instruction and guidance [-@sentance_teaching_2019, p.5]. The approach also responds to calls for computing education to align more closely with sociocultural perspectives [@tenenberg_out_2014], drawing on concepts of mediation and the zone of proximal development (ZPD) [@sentance_teaching_2019, p.2].

PRIMM (Predict, Run, Investigate, Modify, and Make) expands upon the UMC model by incorporating additional learning stages, advocating that learners first predict the outcome of a given code example before testing their predictions against the actual results of running the code. Following this, guided exploration allows learners to investigate possible code modifications before making adjustments to the provided code. The final Make stage, which corresponds to Create in UMC, supports students in developing complete programs or constructing larger components of code from scratch.

The adoption and discussion of PRIMM in learning contexts has grown significantly [@martin_extending_2020; @parry_investigating_2020; @ofsted_research_2022; @barefoot_computing_crystal_nodate], likely due to its concrete applicability within classroom settings and its inclusion in various pedagogical frameworks promoted by the National Centre for Computer Education (NCCE) through computing pedagogy resources [@ncce_using_2020]. However, while PRIMM and LOA pedagogies are not inherently limited to formal schooling, little or no research has yet explored their application in non-school settings. Sentence et al. [@sentance_teaching_2019] acknowledge further limitations, proposing future research to explore co-production possibilities for differentiation in collaboration with teachers, framed within sociocultural approaches. They also suggest more precise alignment between session planning using LOA concepts and lesson delivery incorporating the PRIMM approach.

Social approaches and cultural programmes

The review of CGD&P conducted by Denner and colleagues [-@denner_does_2019] identified social interaction as the third pedagogical strand, observed in 35 of the 68 studies examined. The review, similar to Earp’s [-@earp_game_2015] earlier analysis, highlights limitations in the scope of social and collaborative interventions, which were largely restricted to pair programming approaches and general feedback mechanisms on game projects.

Pair programming is a widely used industry practice that has also been integrated into educational contexts [@hanks_pair_2011]. Pair programming pairs students together, assigning them two distinct coding roles. One student actively codes, while the other focuses on the overall design and logic of the programme. A key advantage of pair programming is its ability to build coding confidence, as students gain experience in both roles. To support novice coders, teachers should model and break down coding processes to improve accessibility.

Werner and colleagues [@werner_pair_2009; @denner_computer_2007] examine pair programming as a strategy to address gender disparities, extending research on collaborative problem-solving in computer programming. They cite studies challenging the gendered dichotomy between bricolage and abstract problem-solving, but highlight the need for further exploration of programming styles and strategies [@denner_computer_2007]. Their research suggests that while pair programming is broadly beneficial, it is particularly effective in narrowing participation gaps related to gender and socio-economic background [@werner_pair_2009, p.31]. In Denner and Werner’s research, pair programming incorporates social learning elements, providing students with greater autonomy in problem-solving strategies and opportunities to construct their identities as programmers. The process of building an identity within a coding community, facilitated through peer collaboration, is central to a socio-cultural understanding of how learners develop programming skills in both classroom and non-formal settings.
The value of these approaches is not confined to pair work and can extend to collaborative groups of larger sizes. For example, the work of Werner and Denner builds upon existing research into collaborative social reality and joint problem-solving spaces, using these frameworks to scaffold the process of ideation [@omalley_construction_1995], while the role of friendly relations is identified as a factor in fostering productive and sustained interactions [@mcdowell_pair_2006]. Additionally, Waite and Sentance [-@waite_teaching_2021] examine the potential for peer instruction in computing education, both as a distinct pedagogy 11 and as a broader collaborative approach to learning programming.

Popat and Starkey [-@popat_learning_2019-1] highlight that collaboration and the sharing of code, as a way of supporting peers, often emerges organically rather than as a deliberately planned aspect of learning designs. Earp and colleagues [-@earp_learner_2013] point to a surprising gap in studies investigating collaboration as both a learning process and an outcome, despite the diverse roles involved in game-making being well-suited to such activity. These gaps are particularly striking given that the development of learner identity within a collaborative culture is widely recognised as fundamental to exploratory, project-based approaches [@kolodner_problem-based_2003].

Game Competitions and Coding Clubs

Coding clubs have been discussed above in relation to the constructionist strand of research, drawing on the early legacy of computer clubhouse initiatives [@peppler_computer_2009]. In a UK context, despite extensive volunteer-led activity, research remains limited on three closely related coding club projects: Code Club, Coder Dojo, and Raspberry Jam 12. Existing research covers various aspects, including the use of Code Club resources by volunteers [@aivaloglou_how_2019], general facilitation approaches, and the absence of an overarching conceptual model at Coder Dojos [@alsheaibi_teaching_2020]. One strategy identified to focus participant activities is the encouragement of competition entries in computing and engineering fields [@aivaloglou_how_2019].

Since 2010, a variety of coordinated online CGD&P competitions have been organised by entrepreneurial organisations, such as the Globey Award from Globaloria, as well as by foundations and public bodies, including Games for Change and the STEM Video Game Design Challenge, sponsored by the White House in the US [@kafai_social_2013]. Game projects also play a significant role in the Collab Challenge, an initiative within the Scratch online community [@kafai_motivating_2014; @kafai2011collaboration]. In the UK, the Coolest Project competition operates in partnership with Coder Dojos and Code Clubs.

Research by Quinlan and colleagues [@quinlan_how_2018; @quinlan2020ideas] examines the Coolest Project, identifying key themes from case study interviews with participants. The study reveals diverse entry points for project engagement, with some students initially drawn to technology, others motivated by a specific idea or problem, and some focused on developing particular skills. The authors also highlight the broader benefits of such competitions, echoing findings from wider research, including the inspirational and motivational impact of seeing peers’ work [@kafai_what_2012; @kafai_motivating_2014], the development of confidence across multiple areas, the enhancement of teamwork and collaborative problem-solving skills, and the significant role of adults in facilitating connections to professional communities and providing logistical support, including parental involvement [@quinlan_how_2018]. This final consideration informs critiques of competition models concerning inequality of access, as not all parents are able to provide this level of support for their children [@thumlert2018learning]. Further concerns arise from the largely uncritical embrace of computing competition values [@thumlert2018learning], a challenge also noted within the broader maker movement [@vossoughi_making_2016].

Educational game jams

A game jam is an event and process characterised by accelerated production method, group collaborations and an ethos of innovation [@arya_international_2013; @gabler2005prototype]. In game jams participants create games in teams (or sometimes individually) in a time-constrained period, typically 24 or 48 hours. While the premise game jams is to promote collaboration, often involving creative constraints to do with the subject of the games to be created, events are inconsistent in their support and scaffolding of collaborative approaches [@goddard_playful_2014]. Team events often take place in physical venues and may be part of a wider global jams [@arya_international_2013]. Meriläinen [-@merilainen_first-timer_2019] notes the potential of game jams but also lack of research on learning mechanism at play in jam events. Aurava and colleagues [-@aurava_game_2021] build on this work with research exploring the use of game jams in formal education contexts, confirming the inherent potential, but also noting the need for outside tutoring and links with existing game jam communities. Arya and colleagues [-@arya_ggj-next_2019] note the limits of the Global Game Jam in terms of the participation of young people and outline the development of the more supported approach of the Global Game Jam Next (GGJN) and a youth programme launched in 2018 targetting the next generation of game designers 13.

GGJN resources are shaped by the educational process developed within the Moveable Game Jam (MGJ) programme [@games_for_change_get_2017], which was created by a collective of New York educators and educational foundations 14. Their game jam guide employs playful methods to enhance inclusivity in the process. To address concerns about inclusivity in adult game jams, various format adaptations have been made. MGJ can be implemented within a shorter timeframe, emphasises low-cost approaches, and supports both digital and analogue offline game production. It features loosely structured activities and broad goals, allowing for significant learner agency. The MGJ process communicates fundamental concepts of game design, particularly through a simplified analysis of game elements, an approach adopted by other game jams and competitions 15 [@games_for_change_get_2017]. Similarly, other key techniques incorporated into later game jam programmes include the use of non-digital games, periodic facilitation, extensive peer testing of games (known as playtesting), adopting playful roles within game creation, and engaging with professional game designers and communities [@fowler_there_2023; @kultima_pikku_2024].

Kultima and colleagues [-@fowler_there_2023] examine the challenges of game jam processes, even with adaptations designed for educational settings. These include the importance of careful planning and clear expectations to align the experiences of novice and more experienced game designers, as well as the difficulties of time management in more formal learning environments. Additionally, Eberhardt [-@eberhardt_no_2016] identifies tensions related to the commercialisation of game jams.

In summary, while research on game jams in educational contexts explores their pedagogical potential, including elements of social capacity building, tangible outcomes, engagement, play, and soft skills, much of the current literature remains focused on feasibility and practical implementation rather than offering a detailed examination of pedagogical processes.

Fifth Dimension interventions & Connected Learning approaches

While this chapter has explored various promising pedagogies and studies aimed at enhancing CGD&P, a common limitation identified across these approaches is the lack of alignment with theoretical perspectives that address social and cultural aspects of learning. To address this, this section examines a series of educational partnerships known as 5th Dimensions (5thD), which were implemented in technology-rich, non-formal settings, often within after-school clubs in lower-income communities [@cole_design-based_2016]. 5thD interventions were designed with a primary research objective of developing sociocultural understandings of learning processes. Initial iterations, led by Michael Cole, took place at San Diego University, supported by volunteers, equipment, and technical assistance to design and deliver a creative series of computer-based, playful activities. Cole [@cole2009designing] describes two contextual motivations behind the design of 5thD: the need for accessible after-school programmes, and the opportunity to provide undergraduate students with practical experience, helping them connect their academic understanding of child development with real-world applications.

The use of novel computer communication technologies such as games and email served a dual purpose, providing opportunities to address reading deficits [@cole2009designing; @cole2014designing] and countering the potential alienation of women and girls from STEM subjects [@cole1987contextual]. 5thD researchers examined the resulting programme as a collaborative initiative, designed to facilitate a joint learning experience between participants and volunteers, each with shared but distinct objectives. The intervention integrated a fictional narrative involving a wizard, with whom young participants engaged via email, motivating them to participate in both digital and play-based activities that simultaneously developed their written and computer literacy skills [@brown_cultural_2008]. The long-term nature of the project led to regular young attendees mentoring new cohorts of student volunteers, supporting their understanding of both the technical and social dynamics of the programme. This process helped young participants develop a sense of expertise, while providing student mentors with valuable insights into the mutual learning processes involved [@cole2009designing].

The importance of designing interventions to support the formation of participants’ identities within the evolving cultures of learning sites became a central aspect of ongoing research [@cole1987contextual] 16. It is significant that varied cultural practices emerged across different settings, responding to the interests and needs of specific learning environments [@valsiner_cultural-historical_2007].

This strand of research into culture formation was further developed by Kris Gutiérrez, who led two 5thD interventions, Las Redes [@scott_nixon_digital_2012] and El Pueblo Mágico [@gutierrez_learning_2019-1]. In both sites, researchers examined the value of a multi-lingual cultural environment in shaping site-specific learning cultures. In related work, Gutiérrez and Digiacomo [-@gutierrez_developing_2008; -@digiacomo_seven_2017] identify the crucial role of learning designers in facilitating identity transitions between different settings, using responsive learning design that enables learners to draw upon funds of knowledge [@moll_funds_1992; @moll1998turning]. Gutiérrez and colleagues [-@gutierrez_learning_2019-1] also clarified an implicit design motivation underlying previous 5thD interventions, noting that the experience should be enjoyable for young people. Fun was essential not only to maintain engagement, but also to foster greater expressions of competence and agency, allowing learners to draw upon their prior experience of play across different contexts 17.

Some of the ideas developed within the 5thD programme were further expanded through their incorporation into an approach called connected learning, part of a project to which Gutiérrez contributed—the Connected Learning Research Network (CLRN) [@ito2013connected]. CLRN examined education for the digital age, drawing on the foundational ethnographic work of Gee and Ito discussed in Chapter 1. It proposed specific principles to guide a broader model of connected learning, focusing on movement between digital and non-digital learning spaces. These principles included ensuring that learning is socially embedded (through peers and communities), interest-driven, and oriented toward opportunity (such as pathways to educational advancement, career success, and civic engagement). The project advocated for expanding access to digital media-related learning opportunities, particularly for under-served youth. This research, through case studies, reports on pedagogical strategies, and approaches to leveraging digital media, has made a valuable and influential contribution to the field. However, while this work includes one case study examining game design using a non-coding tool [@rafalow_connected_2014], it does not provide specific guidance on how to operationalise the broader principles of connected learning as a distinct pedagogy for supporting CGD&P in formal or informal settings.

Design patterns

The guidance for facilitators included within game jam programmes explored earlier incorporates structured analysis of key game design elements. A related strand of research applies a similar approach through the concept of design patterns. Design patterns offer structured solutions to recurring design challenges, illustrated through concrete examples of design principles in context [@alexander1977pattern]. Documenting and sharing design patterns supports the development of design communities, fostering collaborative problem-solving and innovation. The use of design patterns has been widely applied in computer programming and software design [@gamma_design_1995]. In higher education, a design pattern approach is frequently used to teach object-oriented computing, providing scaffolding by modelling community-based design decisions [@fojtik_design_2014-2]. This helps to break down complex problems, making them more modular and concrete [@muller_almost_2004-1; @waite_teaching_2021].

While the modular nature of the knowledge generated supports both replicability and generalisation in research, Dearden [@dearden_pattern_2006, p. 20] advises careful consideration when defining the scope of pattern formulation, noting that patterns that are too abstract may be impractical for real design use, while overly specific patterns may be difficult to adapt to new scenarios. Similarly, Eriksson et al. [-@eriksson_using_2019] expand upon the work of Höök and Löwgren [-@hook_strong_2012] to position design patterns in games as a form of intermediate-level knowledge, bridging the detailed implementation of specific elements with broader theoretical concepts.

The use of design patterns has been widely adopted by game designers and educators working with games in varied ways [@bjork_patterns_2005]. In the professional context of game programming, collections of structural game design patterns serve to share coding practices and establish a common language of game design [@bjork_games_2006]. The term game design patterns (GDP) is applied in different ways, with Kreimeier distinguishing between content patterns and structural software engineering patterns [@kreimeier_case_2002]. Content patterns describe recurring design elements that are visible to the end user. Building on the work of Bergström et al. [@bergstrom_exploring_2010], this study adopts the term GDP to refer specifically to gameplay design patterns, a subset of game design patterns focused on content elements that players can recognise or directly experience as features of the game.

Collections of game design patterns have been used as pedagogical tools in research, helping novices engage with game design processes. In an educational intervention working with 11- to 12-year-olds, Eriksson and colleagues [@eriksson_using_2019] applied a curated selection of patterns to guide learners in analysing and proposing modifications to an existing collaborative game, Stringforce. The study incorporated learner-led game analysis, level design adjustments via a graphical (non-code-based) editor, and co-design of conceptual changes to existing games. Their research built upon a collection of over 200 GDPs, developed in a related research project aimed at supporting novice adult game designers, available as a public resource [@bjork_patterns_2005] 18. The structuring of these patterns drew from broader game theory principles, particularly the MDA framework, which categorises GDPs into game mechanics, dynamics, and aesthetics [@bergstrom_exploring_2010]. Game mechanics refer to the rules and systems that govern gameplay, game dynamics describe emergent behaviours resulting from these mechanics in player interaction, and game aesthetics invoke emotional responses through gameplay elements such as graphics, story, and sound design.

The process of curating patterns for Stringforce involved selecting only 14 from the full collection [@eriksson_using_2019]. The criteria for choosing patterns to include in the co-design stages considered several factors. Concrete patterns were prioritised over more abstract ones to support learner comprehension, selected patterns aligned with the learners’ capabilities, and patterns related to game mechanics were favoured, along with those suggested by the learners. The Stringforce study highlighted several aspects of GDP utility. GDP concepts functioned as a common language between researchers and participants, helped structure analytical framing, and served as reference points for participants’ design proposals [@eriksson_using_2019]. Overall, the findings primarily emphasise benefits for researchers, advancing GDPs as a form of intermediate knowledge that contributes to the field of child-computer interaction research (CCI).

Game Star Mechanic and Scalable Game Design

This section examines Game Star Mechanic and Scalable Game Design, two well-documented educational programmes that incorporate GDPs. Game Star Mechanic (GSM) is an online platform rooted in sociocultural perspectives on learning, designed to foster systems thinking, design thinking, and media literacy through games as multimodal texts [@salen_gaming_2007]. In GSM, video game creation uses a simplified block-based system to modify existing games, eliminating the need for programming [@games_design_2008]. The design choices withing GSM were informed by research highlighting the advantages of using games over other forms of media projects, particularly their interactivity and rules-based structures, which support the exploration of systems thinking and design approaches [@games_gamestar_2010; @tekinbas_gaming_2014; @torres_learning_2009].

The pedagogy embedded in GSM software draws on game theory developed by the Institute of Play founders [@salen_gaming_2007]. This work also influenced the process behind the Moveable Game Jam described earlier. The guiding pedagogy is reflected in an extensive support pack for teachers 19, which incorporates themed challenges focused on categorising key game elements specifically: space, components, mechanics, goals, and rules. In evaluative reviews, Games [-@games_design_2008; @games_gamestar_2010] explores how the design of GSM enables learners to leverage their prior knowledge of game design patterns, fostering complex collaborative design and discourse among novice designers. Games also examines the role of community aspects within the pedagogy, showing how these elements help participants develop an awareness of audience and engage in active critical roles within a design community [@gee_situated_2004].

Scalable Game Design (SGD) is a computing education programme developed and delivered by Colorado University, which, like GSM, created a software tool called AgentSheets to facilitate game creation while supporting teaching resources and training. AgentSheets uses a block-based, drag-and-drop programming approach. Extensive partnership work led by Repenning and Basawapatna [-@repenning_scalable_2010; -@basawapatna_using_2010] enabled large-scale data collection from thousands of school students.

The researchers use the term Computational Thinking Patterns (CPTs) to describe patterns present in computer games that they support learners in coding. The familiarity and clear applicability of these patterns to specific learning outcomes guided SCG researchers towards a pattern-based approach, in contrast to more abstract interpretations of computational thinking. Their interactions with teachers highlighted the importance of foregrounding concepts with potential applications in science simulations [@basawapatna_recognizing_2011].

The authors provide examples of computational thinking patterns, such as generation and absorption in predator-prey relationships,

Another important concept within the SGD pedagogy is student ownership, particularly in participants’ ability to design their own characters and backgrounds [@repenning_scalable_2015]. One limitation of the approach is the reliance on heavy scaffolding, requiring step-by-step instructions due to the complexity of the game authoring process [@repenning_scalable_2015, p.11.10]. While the authors suggest allowing students to create their own games, this is presented as an optional activity at the end of a unit, which many schools could not implement due to time constraints. This may explain the relatively low figures for student ownership, with only 139 out of 700 responses (20%) indicating aspects of student ownership [@repenning_scalable_2015, p.11.10]. Other limitations include the programme’s reliance on bespoke software developed by the team, which raises concerns about long-term maintenance. Additionally, while the resources are described as just-in-time instruction, a review of the supporting website reveals lengthy instructional materials, which may hinder their effectiveness in achieving this goal 20.

Conclusion

Chapter 1 outlined the study’s key needs, framing the problem through barriers to participation, including technical challenges, guided pedagogies, and cultural factors. This chapter has examined promising pedagogical approaches in CGD&P research and digital making more broadly, while also identifying gaps that require further investigation. This section revisits the research questions and objectives to refine the problem statement of this thesis.

Regarding RQ1 21 and the research gaps in structural and technical approaches, while various game-making tools exist, my focus on authentic tool usage and community engagement has led me to select a toolset comprising a code playground and a professional JavaScript game library. This approach has received little attention in related research, presenting an opportunity for further exploration. Similarly, while UMC [@lee_computational_2011], half-baked games [@kynigos_children_2018], and PRIMM [@sentance_primm_2017] demonstrate clear potential in this area, this study proposes further development by integrating sociocultural pedagogies outlined in the connected learning and 5thD programmes. Additionally, as few CGD&P studies convey pedagogical detail in a way that facilitates replication, this study addresses this limitation by carefully documenting learning processes and publishing facilitation materials as OERs.

Turning to research gaps in the potential of GDPs, RQ2 directly addresses this issue 22. The use of design patterns shows promise in supporting motivation and structuring documentation [@repenning_scalable_2015] as well as guiding feedback processes within co-design methodologies [@eriksson_using_2019]. However, existing studies present certain limitations. In the SGD programme [@repenning_scalable_2015], constraints in participant pathways contributed to reduced student ownership. In the Stringforce study, the potential applications of GDPs for both participants and teachers remain underexplored [@eriksson_using_2019]. Additionally, as participants did not engage with computer code, further research is needed to assess how a collection of GDPs might enhance programming, particularly in addressing the ongoing challenge of navigating abstract and concrete dimensions within the learning process.

Addressing agency and cultural identity within programming environments for novices in relation to RQ3 23, this literature review identifies a need for further research on pedagogical approaches that support the formation of game-making identities and self-expression [@denner_does_2019; @kafai_constructionist_2015; @kafai_connected_2016]. While various concepts related to participant empowerment are present in relevant research, the ideas of fluency and agency, as well as the processes that shape and facilitate their development, remain under-theorised within CGD&P.

Thus, while some existing pedagogies are suitable and inclusive, there remains a need for more approaches that are specific, aligned with non-formal spaces, and support agency development and learner identity formation by drawing on funds of knowledge. Having examined the gaps within the relevant body of literature, it is clear that a research approach enabling detailed exploration of context, tool use, guiding pedagogical processes, and learner agency is necessary.

In the following chapter, I outline the study’s approach to achieving these objectives, drawing upon activity theory as the principal theoretical framework.

  1. “Pedagogy refers to that set of instructional techniques and strategies which enable learning to take place and provide opportunities for the acquisition of knowledge, skills, attitudes and dispositions within a particular social and material context.” [@siraj-blatchford_researching_2002] ↩

  2. LOGO was a simplified text based computer language developed to help novice coders. ↩

  3. Available at https://web.archive.org/web/20220901000000*/https://www.familycreativelearning.org/s/Family-Creative-Learning-Guide-2017.pdf ↩

  4. Stepwise approaches as used in the study refer to both step-by-step instructions and incremental project challenges which increase in difficulty or scope. ↩

  5. See the the work of Denning for an overview of critique of Wing’s computational thinking [@denning_remaining_2017] ↩

  6. Constructivist and sociocultural schools of research which underpin research on PBL are addressed in Chapter 3. ↩

  7. The role of organisations like PBL Works and Edutopia have been significant in sharing accessible, educator focused support for teachers and facilitators. The design rubric from PBL Works mentioned is available here. https://my.pblworks.org/resource/document/project_design_rubric ↩

  8. Funds of knowledge are resources which learners can draw on in either formal or informal settings. These may be home practices learned from family members, special hobby interests or broader cultural practices [@moll_funds_1992] ↩

  9. A good example of supporting resources which outline use of resources for teachers and learners is the work around Adventure Author. https://web.archive.org/web/20120511180037/http://judyrobertson.typepad.com/adventure_author/teaching-materials.html ↩

  10. Here the concept of affordances is to express the delegated intention of designers to create features for users that encourage certain desired behaviours [@mcgrenere_affordances_2000]. ↩

  11. Peer instruction is also a specific, perspective technique as well as a general product of collaborative learning environments. [@crouch_peer_2001]. ↩

  12. Grass roots approach characterised by a large take up of enthusiastic community activity in response to a model encouraging others to organise their own events. All three projects have been subsumed into the RPI foundation, which offers mostly instruction based support via their website which disappointingly may well indicate a lost opportunity to study innovative practice. ↩

  13. Global Game Jam is one of the largest, international game jam programmes aimed at adults. Global Game Jam Next is an educational sub-project with supporting resources for facilitators. https://ggjnext.org/curriculum/. From 2025 has partnered with Games 4 Change (G4C) which also has a game jam guides for their student challenge https://gamesforchange.org/studentchallenge/ ↩

  14. The emergence of the experimental process between NYC Hive and other partners and individuals is documented via a blog post on the games for change website. https://web.archive.org/web/20180212051341/https://gamesforchange.org/studentchallenge/2016/11/21/10569/ ↩

  15. The elements are simplified into space, goal, components, mechanics and rules, a framework adopted by GGJN and G4C These terms are explored in Chapter 5. ↩

  16. Idioculture is a term used to describe these micro-cultures by Cole and other 5thD researchers. It is explored in Chapter 3. ↩

  17. Gutiérrez [-@gutierrez_when_2020-1; -@gutierrez2014integrative] has developed the concept of horizontal movement of practice between between sites of learning rather than a vertical top-down transmission model of learning in ways that are discussed in Chapter 3 within the theoretical framework of this thesis. ↩

  18. A collection of over 200 patterns organised by diverse themes are online here. http://virt10.itu.chalmers.se/index.php/Category:Patterns ↩

  19. A comprehensive guide has been created for GSM aimed at practitioners supporting the use of software with structured sessions. https://web.archive.org/web/20131220180134/https://sites.google.com/a/elinemedia.com/gsmlearningguide/home ↩

  20. An example of a worksheet created is available from the supporting resource site for the Scalable Game Design programme. https://web.archive.org/web/20220331110501/https://wiki.computationalthinkingfoundation.org/wiki/images/5/5b/ACO_Frogger_Student_v1.0.pdf ↩

  21. RQ1: What contradictions emerged during participation in CGD&P activities and how were they addressed via an innovative pedagogy? ↩

  22. RQ2: How can the use of a collection of game design patterns support CGD&P, in particular in relation to abstract and concrete dimensions of existing pedagogies? ↩

  23. RQ3: How do learner agency and game-maker identity develop within CGD&P communities of practice, and what pedagogical strategies best support this evolution across diverse learning contexts? ↩