Keywords

1 Introduction

The use of educational games for learning, known as Digital Game-Based Learning (DGBL), has increased over the last few years [27]. Some of the reasons are that educational games not only serve to entertain but also contribute to facilitate interactivity during learning, overcome the lack of commitment on the part of students, provide opportunities to think and reflect deeply and perform a positive change in behaviour [15].

Educational games development process follows the classical game development process [5, 25]. In pre-production phase, the game is designed according to educators’ requirements, which define learning outcomes, interactive learning activities, and their evaluation metrics. In the production phase, the game is constructed by game developers, and validated by teachers through an iterative and incremental process. Finally, in the distribution phase, support and training are carried out accompanied by instructional designers. Nevertheless, teachers are usually dependent on developers to update or modify the game to better fit learners and learning outcomes of each session or academic unity. As a consequence, educational games may have short-life usage in the classroom. To overcome this limitation, several educational games platforms offer authoring tools by mean of scripting, visual editors and mods components, with the drawback of being intended for users with technical, artistic or programming skills.

Authoring tools have been successfully proposed to assist the development of a game in different ways, either from scratch or by means of editing already generated content and game mechanics, abstracting the technical aspects of game development for non-programmer users.

There are different approximations to educational game authoring. First, successful commercial games, such as Minecraft: Education Edition Footnote 1 and SimCityEdu Footnote 2, allow teachers create new scenarios or customise pre-existing configuration levels to learners using embedded educational mods.

Second, other tools let the edition of the concrete game features on pre-built games such as scenarios, characters, and dialogues. SeGAE is a serious game authoring environment that offers instructors the options to modify the game design by defining rules and game objects such as characters, objectives, victory conditions and authorised actions even after the development stage [28]. SHAI Scenario Editor allows instructors design educational scenarios from high-level logic diagrams to low-level logic structures [26]. e-Adventure is an educational game authoring tool that adds the possibility of exporting the game to an LMS (Learning Management System) [24].

Finally, there are approaches that propose tools for building games from scratch, incorporating both learning and gaming aspects from early stages in the development process. Roungas and Dalpiaz [19] developed a web-based environment that assists game designers in the creation of a game design document based on a conceptual model for educational games. Kirkley et al. [16] propose a wizard tool that allows both instructional designers and game designers to link instructional design elements and game events. Tang et al. [23] developed a model-driven serious game development platform supported by a game content model [22] in the design specification. StoryTec uses a visual programming approach to allow teachers without programming skills to specify generic games activities (e.g. puzzle, quizzes) applicable to any discipline. Then, teachers have to define concrete instructional content for those activities, according to the Intended Learning Outcomes (ILOs) [17]. As this process of defining instructional content for the game can be arduous and time-consuming, an alternative authoring approach can be based on embedded specific curricular activities. By embedded specific curricular activities, we mean digital teaching and learning activities (TLAs) integrated into the game and directly related to a concrete content of a discipline, such as fractions in maths, human muscles in natural sciences, etc. Therefore, the authoring tool allows teachers configure some properties of the TLAs. For example, in a fractions game, the teacher can specify the values of denominators that he wants students to practice in the game challenges.

In this work, we propose a conceptual model that integrates both game and educational concepts with the purpose of guiding the development of an educational authoring tool. The model is defined to cover two different authoring domains (educational and game domains). In the educational domain, the authoring consists in create/edit either generic activities of any discipline (i.e. puzzle mechanics to be applied to reach different ILOs) or embedded specific curricular activities (already designed activities, integrated into the game, to achieve concrete ILOs). In the game domain, the authoring focuses on create/edit game elements and activities, e.g. challenges, characters, levels, etc. Additionally, and as a proof of concept, we show two authoring applications of this conceptual model in the domain of maths.

This paper is structured as follows. Section 2 presents a proposal of the educational game authoring process. Section 3 presents our conceptual domain model that supports instructional designers in game authoring taking into account the Outcome-Based Education (OBE) approach. Section 4 illustrates the applicability of our proposal through a serious game authoring platform about mathematics concepts. Finally, Sect. 5 gives conclusions and future work.

2 Educational Game Authoring Development Process

In the process of educational game authoring development, we should consider both the educational and the game domains to identify the elements of each domain susceptible of being authored. To identify these elements, we pass through the classical development process of a digital educational serious game, which has got three phases: pre-production or planning phase, production or construction phase, and distribution or delivery phase [5, 15].

In the pre-production phase, the game concept is developed. In this phase, from the point of view of the educational domain, instructional designers define elements such as Intended Learning Outcomes (ILOs), their metrics, Learning Styles (LS), Teaching and Learning Activities (TLAs) to be considered and how to assess the learning [14]. From the game domain viewpoint, designers define elements including the concept, flow, game objectives related to learning outcomes, and players information such us: age, gender, preferences and player types. In this phase, game design starts defining the genre, goals, some aesthetics, mechanics, dynamics, number of players, the input device, feedback elements and description about the game impact. At this stage, it should be defined the domains of authoring (educational and/or game domains), and the elements of these domains to be authored.

In the production phase, the design, programming, art and quality assurance of the digital serious game is carried out taking into account the authoring domains defined in the previous stage. Additionally, and in parallel, the authoring tool is designed and developed in the specified authoring domain.

Finally, in the distribution phase, diffusion management, support and training are performed accompanied with instructional designers. At that phase, the validation of the proposed learning assessment is a key factor to get an early feedback to adapt either the gameplay or the learning activities included in subsequent iterations of the development. In this phase, we propose that, once the game is finished and the authoring tool is deployed, the instructor uses it to tune some game parameters (i.e. in a game about geometry topics, the teacher can configure the inclusion of embedded learning competences in the game challenges such as identification of 2D shapes and their properties, and he can specify the proportion of activities for each ILO). Additionally, it is possible to include either personalised activities for each student or activities of specific ILOs for the entire class, to suit teacher’s own pace.

Phases in the afore-described process should follow principles of User Centered Design (UCD), with a strong focus on users during all the steps of the development, and with the aim of evaluating as early as possible prototypes to get fast and valuable feedback from users.

As the target group of an educational authoring tool are teachers, usability criteria such as easy to learn, error prone, efficacy and efficiency are critical to design a product that fits their needs and values.

The design of the educational game, with a target group of students aiming at both learning and especially having fun, takes a broader perspective. It also focuses on usability criteria but takes into account hedonic factors closely related to the experience of the user (UX) in the game, such as game play as well as immersion, and emotions. In the literature, the research on these (and other) factors provides heuristics to assist game design development [9].

3 A Conceptual Model for Educational Game Authoring

Our proposal is a conceptual domain model for educational game authoring based on the Outcome-Based Education (OBE) approach, which supports educators in the elaboration of digital instructional design [20].

Our conceptual domain model for educational game authoring presents three main components. Figure 1 depicts a simplified version of the model that will be further detailed in the following subsections: The OBE Educational Model (OEM), that considers the instructional design; the Game Design Model (GDM), that addresses the digital game-based design; and the Digital Game-Based Learning Model (DGBLM), that defines the Digital Game-Based Learning (DGBL) approach supported by both OEM and GDM.

Fig. 1.
figure 1

Main conceptual model components for supporting educational game authoring.

Full size image

3.1 OBE Educational Model (OEM)

The Outcome-Based Education (OBE) is an educational approach that focuses on the process of learning followed by each student. OBE proposes three guided stages that identify the key elements of the deployment of the educational curriculum: the Intended Learning Outcomes (ILOs), the Teaching and Learning Activities (TLAs) to lead to the ILOs, and how the students are assessed [6].

In the first stage, the educator designs the syllabus and classes of an outcomes-based curriculum regarding the learner’s accomplishment [3, 21]. Therefore, ILOs describe the competencies that students must achieve as well as teachers must be able to measure students’ learning achievement. Then, at this stage, instructional designers ought to define metrics to monitor student learning during the ongoing process (formative assessment), and also to evaluate student learning at the end of the process (summative assessment) [13].

In the second stage, the educator defines Teaching and Learning Activities (TLAs) to achieve the ILOs taking into account the learning styles (LS) of students. In the cognitive domain, TLAs can be deployed using different categories of the revised Bloom’s taxonomy [1]. Each category involves the use of its own specific Learning Mechanics (LMs) [2, 4].

Finally, the third stage focuses on the deployment of the TLAs and the learning assessment which implies collecting, analysing, interpreting and using information to improve students’ learning ongoing process [11].

Model in Fig. 2 depicts entities that support the instructional design process in the OBE educational approach, and takes into account well-known methodologies used in teaching: ADDIE Model [18], the Dick and Carey Systems Approach Model [10], the Gagné’s 9 Events of Instruction [12] and Bloom’s Taxonomy [7]).

Fig. 2.
figure 2

OBE Educational Model (OEM).

Full size image

The instructional designer starts the process of elaborating the Instructional Design for a Course considering Topics, Intended Learning Outcomes in knowledge, skills and attitudes that learners have to achieve, Learning Objectives, analysis of Learners regarding Learning Styles and Learner Profile, Learning Environment, designing Formative and Summative Assessment with Criteria, Method and Feedback, and designing Teaching and Learning Activities (TLAs) with their Learning Mechanics (LM), Metrics. The Instructional Design is adapted by the Instructor to derives a Syllabus and uses Learning Materials.

Some of the just mentioned concepts are candidates to be configured by teachers and so be incorporated in an authoring tool. For example, TLAs are closely linked to game challenges and game mechanics in the game domain, and can be as generic such as puzzles or quizzes, or domain-specific (i.e. dependent on the learning subject) such as construct a 3D building using basic shapes.

3.2 Game Design Model (GDM)

This model contains entities that define main components in the game design process, such as the Game, analysis of Game Requirements, Game Players, their Player Types and Game Controls, outline the Game Challenges, makes a Game Narrative and Game Presentation, define the Artistic Style, Characters and Game Objects, design the User Interface, Environment, Levels with Music, Maps, Activities with GamePlay, GameRules, Game Mechanics and Progression (See Fig. 3).

Some of these game entities are susceptible of being configured by the teacher and then can be incorporated in the authoring tool. For example, a teacher can configure the time to finish a Game Challenge.

Fig. 3.
figure 3

Game Design Model (GDM).

Full size image

3.3 Digital Game-Based Learning Model (DGBLM)

This model contains entities that integrate elements of the OBE Educational Model (OEM) and Game Design Model (GDM). It focuses on relevant components which can be considered by authoring tools.

When instructional designers or instructors assume the role of game author, they take into account Digital Instructional Components, Digital Guidances, Digital Materials and Technology, as well as Context Requirements to define the Digital Instructional Design, which relies on Digital Educational Games (DEG) including DEG Activities according to teaching and learning activities (TLAs), Learning Mechanics (LM) and Game Mechanics (GM).

The learners are also players (DEG Learner Player) and they have a DEG Personal Activity Portfolio about performance in academic and game progression, the DEG Learners/Players can be grouped according to their characteristics (DEG Learner Player Group). The DEG Outcomes Assessment Design generates a DEG Assessment supported by results of DEG activities following a DEG Rubric giving DEG Feedback both in the game and academic progression. The Digital Educational Games (DEG) included in the Digital Instructional Design can have different DEG Activities with DEG Activity Types based on learning styles and player types, hence, different DEG GamePlay. According to Kapp [15], the types of game activities for learning games are Building, Role-Playing, Exploring, Strategising, Puzzle, Allocating, Collecting and Matching games (See Fig. 4).

Fig. 4.
figure 4

Digital Game-Based Learning Model (DGBLM).

Full size image

4 Authoring Tools in Practice

In this section, we show how the conceptual authoring model just presented materialises in two math games and their corresponding authoring tools. Both authoring proposals support the edition of specific curricular activities in the educational domain. In the first game, named Fracsland, we put the focus on fractions concept and present the learner an adventure in an island with fractions related challenges. The second game, called GeoPieces, is related to geometry learning both in 2D and 3D. Both games are addressed to 10–12 aged children.

In the following, we refer in italics those concepts of the domain model proposed in the previous section (see Fig. 1) that have been used in Fracsland and GeoPieces authoring tools.

4.1 Fracsland

This game is an adventure serious game where the players assume the role of a shipwreck survivor with the goal of escaping from an island. They collect pieces of a boat engine by interacting with game characters and solving fraction quests.

In this game, the Intended Learning Outcome (ILO) is to be able to understand and apply the fraction concept in practice. To cover this ILO, teachers usually define a TLA (Teaching and Learning Activity) as a puzzle based on the manipulation of materials (e.g. wooden sticks, pies) and also use the blackboard [8].

This puzzle is mapped in the educational game as a challenge the player has to complete interacting with “fractionable interactive objects” (e.g. bridge) using an inventory of tools shown in the HUD (See Fig. 5).

Fig. 5.
figure 5

Screenshots of fractions’ challenges in Fracsland: an adventure serious game in maths.

Full size image

The HUD includes resources collected during the game (such as rope, wood, knife) and fractions tools (2/9, 6/9, ...). These tools can be used to select those parts of the resources that are needed to solve challenges (e.g. use a part of the wood to complete a bridge in construction).

The parameter of the challenge that will be configurable by the teacher is the denominator of the fraction that the educator wants their pupils to work with. Eventually, the teacher can define the solution of the quest (i.e. both numerator and denominator). Another parameter that the teacher can define is how to represent the fractions in the game (pie or bars).

As shown in Fig. 6, Fracsland authoring platform allows configuring the DEG Learner/Player Group. In this case, the teacher can upload either an excel file with all the class or add each student manually.

Fig. 6.
figure 6

Fracsland authoring tool: DEG learner/player group.

Full size image
Fig. 7.
figure 7

Fracsland authoring tool: DEG activities.

Full size image
Fig. 8.
figure 8

Fracsland portfolio.

Full size image

Figure 7 shows three DEG Activities corresponding to three game challenges configured by the teacher. The first DEG Activity shows that the teacher wants their students work with x/2 fractions where x representing any possible numerator in the game challenges. The same for the rest of the challenges, x/5 and x/7. These activities will be included in the game that the teacher can automatically generate for the entire classroom. Additionally, the teacher can personalise quests individually. Figure 8 depicts DEG Personal Activities Portfolio that contains the set of challenges the teacher assigns to one of the students (i.e. named Iker). Figure 9 depicts in green dotted lines, those entities of the proposed conceptual model developed in Fracsland.

Fig. 9.
figure 9

Main conceptual model for supporting educational game authoring. The entities in green and red dotted lines were developed in Fracsland and GeoPieces respectively. (Color figure online)

Full size image

4.2 GeoPieces

The game consists of a series of campaigns that the students will have to finish in order to obtain complex 3D objects from simpler three-dimensional shapes. To gain them, it will be necessary to get first some 2D shapes and then build basic 3D objects through unfolding (See Fig. 10). Each campaign covers a set of ILOs (Intended Learning Outcomes).

Fig. 10.
figure 10

Screenshots of GeoPieces game.

Full size image
Fig. 11.
figure 11

GeoPieces authoring tool.

Full size image

This game is designed to support the following ILOs: identification of 2D shapes and their properties, such as symmetries, angles, perimeters and areas, the identification of 3D objects and their relationship to 2D and 3D shapes. In this game, the authoring tool allows teachers to define the ILOs that are deployed in form of game activities (campaigns) and let assess learner progression, that is, the teacher can say“let my students play only with challenges related to certain learning competences, such as identification of 2D shapes and their properties”. Then after some days, when new knowledge is acquired by the children, the teacher can enable the following learning competences related to 2D and 3D shapes. Alike Fracsland where the teacher configures TLA’s parameters that map directly to game components (i.e. modifying working fractions in the game with new values), GeoPieces authoring hides the details of the game activities from the teacher. Additionally, as a campaign focuses on a set of ILOs, teachers can weight ILOs to be deployed in the game. For example, in a 2D geometry learning session, if the ILOs named “Classification, Angles and Parallelism” were weighted by 40% and “Symmetry, Rotation, and Proportions” ILO by 60%, and the rest of ILOs by 0%, the number of game activities included in the campaign would be proportional to these weights.

As can be seen in the left part of Fig. 11 the teacher can select the campaign (i.e. the 3D object to be constructed) between a set of predefined campaigns, or can select to create a new one using the button “+”. The middle part shows the percentage of each ILO that will be used to generate game activities (i.e. DEG-Activity) for the selected campaign. In the right part of the Figure, teachers can configure some properties of the activity, such as time constraint (DEG-GamePlay), and interactive helpers (e.g. rule, compass), represented as Digital Material. In the bottom right part, the error menu specifies the number of minor and major errors that will trigger a reporting alarm to the teacher (Assessment Results). Figure 9 depicts in red dotted lines, those entities of the proposed conceptual model developed in GeoPieces.

The play list in Youtube “A Conceptual Model for Educational Game Authoring: a Showcase in Math Games”Footnote 3 shows videos about the games Fracsland and GeoPieces.

5 Conclusions and Future Work

In this paper, we propose a conceptual model that supports the development of educational game authoring tools to allow educators customise game activities according to their intended learning outcomes. Although state of the art approaches consider educational aspects, these leave aside key elements about the educational approach such as intended learning outcomes and its metrics, learning styles, teaching and learning activities mapped into the game, game mechanics matched with learning mechanics according to Bloom’s taxonomy in the learning domain, player types, and embedded assessment in digital learning activities. Furthermore, the proposed model is designed to support different domains of authoring: educational (generic and specific curricular learning activities) and game (scenarios, characters, challenges, etc.) domains.

Finally, as a proof of concept of the authoring tool in the educational domain, we present two examples of maths games that can be configured by teachers, either editing game activities’ parameters related to the Teaching Learning Activities (TLAs), or defining the percentage of Intended Learning Outcomes (ILOs) that will guide the generation of number and type of game challenges. Moreover, this generation can be personalised according to students’ profile.

As on-going work, we are further developing the educational game authoring platform based on the proposed conceptual model, and plan to deploy and validate it with teachers in classrooms.