ChapterPDF Available

Personalized Support, Guidance, and Feedback by Embedded Assessment and Reasoning: What We Can Learn from Educational Computer Games

Authors:

Abstract and Figures

Software that intelligently interprets the goals and needs of its users on the basis of their behaviors without interrupting the work flow and consequently disturbing concentration and software that can support the users in a personalized, smart, yet unostentatious way is a desirable vision, for sure. One attempt to such support system was Microsoft’s famous paperclip. The underlying logic, unfortunately, was rather simple and the users did not accept the feature very well. This paper introduces a psychologically and formally sound approach to a non-invasive, hidden assessment of very specific needs of the users as well as their competencies and corresponding tailored support and feedback. The approach was developed in the context of adaptive digital educational games and is based on the concepts of Competence-based Knowledge Space Theory as well as that of Problem Spaces. The purpose of this paper is to broaden the concept and elucidate a possible bridge from computer games to regular software tools. KeywordsEmbedded Assessment-Micro Adaptation-Support Methods-Feedback-User Model
Content may be subject to copyright.
Personalized Support, Guidance, and Feedback by
Embedded Assessment and Reasoning: What We Can
Learn from Educational Computer Games
Michael D. Kickmeier-Rust, Dietrich Albert
To cite this version:
Michael D. Kickmeier-Rust, Dietrich Albert. Personalized Support, Guidance, and Feedback
by Embedded Assessment and Reasoning: What We Can Learn from Educational Computer
Games. Peter Forbrig; Fabio Patern´o; Annelise Mark Pejtersen. Human-Computer Interaction,
332, Springer, pp.142-151, 2010, IFIP Advances in Information and Communication Technology,
978-3-642-15230-6. <10.1007/978-3-642-15231-3 15>.<hal-01055486>
HAL Id: hal-01055486
https://hal.inria.fr/hal-01055486
Submitted on 12 Aug 2014
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destin´ee au d´epˆot et `a la diffusion de documents
scientifiques de niveau recherche, publi´es ou non,
´emanant des ´etablissements d’enseignement et de
recherche fran¸cais ou ´etrangers, des laboratoires
publics ou priv´es.
Personalized Support, Guidance, and Feedback by
Embedded Assessment and Reasoning: What we can
Learn from Educational Computer Games
Michael D. Kickmeier-Rust and Dietrich Albert
Department of Psychology, University of Graz
Brueckenkopfgasse 1, 8020 Graz, Austria
{michael.kickmeier; dietrich.albert}@uni-graz.at
Abstract. Software that intelligently interprets the goals and needs of its users
on the basis of their behaviors without interrupting the work flow and
consequently disturbing concentration and software that can support the users in
a personalized, smart, yet unostentatious way is a desirable vision, for sure. One
attempt to such support system was Microsoft’s famous paperclip. The
underlying logic, unfortunately, was rather simple and the users did not accept
the feature very well. This paper introduces a psychologically and formally
sound approach to a non-invasive, hidden assessment of very specific needs of
the users as well as their competencies and corresponding tailored support and
feedback. The approach was developed in the context of adaptive digital
educational games and is based on the concepts of Competence-based
Knowledge Space Theory as well as that of Problem Spaces. The purpose of
this paper is to broaden the concept and elucidate a possible bridge from
computer games to regular software tools.
Keywords: Embedded Assessment; Micro Adaptation; Support Methods,
Feedback, User Model
1 Introduction
An important trend in the area of learning technologies is game-based learning.
Computer games are a tremendously successful, popular, and attractive genre. A
substantial number of young people spend many hours a week playing computer
games, and these games are often the preferred play. Thus, using the motivational
potential of computer games for educational purposes may open new horizons for
educational technology and is challenging educators and developers. The very nature
of utilising (computer) games for education is that playing games is one of the most
natural forms of learning. Children learn to talk by playing with sounds and learn
collaboration and strategic thinking when playing role playing games such as
Cowboys and Indians. Immersive digital educational games (DEG) fall back to that
origin and are a highly promising approach that makes learning more engaging,
satisfying, inspiring, and maybe even more effective (see [1]). The major strengths of
DEGs are generally observed at a high level of intrinsic motivation to play and
proceed in the game, a meaningful yet rich and appealing learning context, immediate
feedback, and a high level of interactivity, challenge, and competition. Some
researchers even argue that the exposure to “twitch speed” computer games, MTV,
and the Internet has altered cognitive processes, emphasising specific cognitive
aspects while de-emphasising others [2]. Thus, the so-called “digital natives” may
require different, possibly non-conventional, educational approaches. Current
challenges in the area of DEGs are seen in, for example finding an appropriate
balance between gaming and learning activities (e.g., [3]), finding an appropriate
balance between challenges through the game and abilities of the learner (e.g., [4]),
convincingly embedding educational objectives in a game scenario, particularly when
declarative knowledge is concerned (cf. [5]), or managing the extensive costs of
developing high quality games (e.g., [3]).
In closer examination, many of those challenges are very similar to the challenges
of interaction principles and interaction design with “regular” software”. The digital
natives, Marc Prensky [2] was talking about in the context of DEGs, are also users of
regular software products (office tools, web browsers, websites, communication tools,
etc.) and demand very specific – maybe unusual – features. Moreover, we are facing a
large group of users who are highly inexperienced and sceptical towards the use of
computer devices of all kinds, for example the elder generation. Examples of
requirements might be balanced and intuitive multi-tasking possibilities, an embedded
and personalized support with complex functionalities, integrating various tools for
various purposes in common and usable devices, or managing high development and
support costs.
In the focus of the present paper is to address a major challenge for research
in the context of translating ideas from game-based learning to the usability of regular
software products. In DEGs it is all about an individual (learning) experience in order
to reach educational effectiveness and maintain fun, immersion, flow experience1, and
the motivation to play and, therefore, to learn. Thus, meeting individual preferences,
interests, abilities, and goals is key to successful game-based learning. Such claim, of
course, is also a highly desirable asset of regular software. In the following sections
we will present a formal cognitive approach for an non-invasive, embedded
assessment of psycho-pedagogical aspects such s motivational state or learning
progress and corresponding personalized support by subtle feedback and support,
which is currently implemented in learning adventure games.
2 Intelligent Adaptation and Personalization in DEGs
To meet the aforementioned claims, DEGs are supposed autonomously adapt to the
learner along a variety of axes, for example, prior knowledge, learning progress,
motivational states, gaming preferences, and psycho-pedagogical implications.
Generally speaking, adaptive approaches to technology-enhanced learning contest the
one-size-fits-all approach of traditional learning environments and the attempt to
1 According to Mihaly Csikszentmihalyi [6], flow refers to a highly immersed experience when
a person is engaged in a mental and/or physical activity to a level where this person loses
track of time and the outside world and when performance in this activity is optimal.
tailor the learning environment according to individual needs and preferences. The
spectrum of approaches, methods, frameworks, and applications is quite broad [7]
there are basically three major concepts:
Adaptive presentation: adjusting the look and feel of a learning environment
according to individual preferences or needs; for example, different colour
schemes, layouts, or amount of functionality;
Adaptive curriculum sequencing: providing the learner with learning tailored
to individual preferences, goals, learning styles, or prior knowledge;
Adaptive problem solving support: providing the learning with feedback,
hints, or solutions in the course of problem solving processes.
Most of these methods and frameworks for adaptation and personalisation were
developed in the context of conventional e-learning. The underlying concepts and
ideas are currently extended and adjusted to the requirements of the rich virtual
gaming worlds, particularly to maintain an immersive gaming experience and high
levels of motivation, curiosity, and flow experience [8].
A method, which is highly interesting for interaction design in general, is an
approach to non-invasive assessment of knowledge and learning progress in the open
virtual worlds of computer games and a corresponding adaptation by personalised
psycho-pedagogical interventions. The approach, labelled micro adaptivity, was
developed in the context of 80Days (www.eightydays.eu), a multidisciplinary
research and development project. The project had the ambitious goal of utilising the
advantages of computer games and their design fundamentals for educational
purposes and addressing specific disadvantages of game-based learning. Within the
project, a methodology for the successful design of educational games was
established, and a game demonstrator was developed based on a state-of-the-art 3D
adventure game (see Fig. 3 for some impressions of the game).
3 Adaptation on the Micro Level
A primary task for (game-based) adaptive mechanisms is to guide and support the
user in accomplishing specific goals, for example, informing the user, intervening
when misconceptions or errors occur or the work/learning progress is unsatisfactory,
hinting, or providing the user with appropriate feedback. In addition, tasks are
motivating, maintaining immersion, and personalising the environment according to
the preferences and needs of the user. Accomplishing this goal requires a theoretical
and technological approach that enables the system to assess cognitive states (e.g.,
competence states or motivational states), (learning) progress, possible
misconceptions, or undirected/unsuccessful problem solving strategies. A distinct
characteristic of adaptive DEGs is that gathering the necessary information from the
user cannot occur in a conventional form (e.g., by test items, questions, or tasks). A
DEG requires an assessment that does not harm motivation, immersion, flow
experience, or the game’s storyline [8]. The very basis of micro adaptivity is a formal
psychological model for interpreting the behaviour within the virtual environment.
To realise the vision of a non-invasive support of users in DEGs (i.e., embedded
assessment and subsequent educational interventions), we combine the Competence-
based Knowledge Space Theory (CbKST), which is a formal set-theoretic
psychological approach to human learning and development that has been
successfully utilised in conventional adaptive e-learning, with cognitive theories of
problem solving. This theory provides a detailed domain model that includes a set of
meaningful competence states as well as a set of possible learning paths. Problem
solving theories, in turn, provide a set of possible problem solving states as well as
possible problem solving paths. While in an DEG (at least in most types of DEGs)
complex problem solving is an important mechanisms of self-directed, constructivist
learning, for regular software we can assume that the user’s goal is to accomplish a
specific task or to solve a specific problem that, in turn, requires performing a
sequence of specific action and interactions with the software. To give a very simple
example, the task might be to highlight a certain word of a text in an office software.
The steps to solve this problem are (i) taking the mouse and highlighting to word by
setting the cursor at the beginning or the end of the word, pushing the left mouse
button and by moving the mouse with pressed mouse button along the word and ,
finally, by releasing the mouse button. When the word is selected, (ii) the user has to
press the correct highlighting symbol in the tool bar with the left mouse button.
3.1 Competence-based Knowledge Space Theory
In essence, CbKST [9, 10] originates from the Knowledge Space Theory (KST)
established by Jean-Paul Doignon and Jean-Claude Falmagne [11, 12], which is a
well-elaborated formal cognitive framework for covering the relationships between
problems (e.g., tasks or test items). It provides a basis for structuring a domain of
knowledge and for representing the knowledge based on prerequisite relations (Fig.
1, left panel), in the sense that one task is easier than another or that it is likely
mastered before another. While KST focuses only on performance/behaviour (e.g.,
solving a test item), CbKST introduces a separation of observable performance and
latent, unobservable competencies, which determine the performance. Essentially,
CbKST assumes a finite set of competencies and a prerequisite relation between those
competencies. As mentioned, a prerequisite relation states that a competency a (e.g.,
multiplying two positive integers) is a prerequisite to acquiring competency b (e.g.,
dividing two positive integers); if a person has b, we can assume that the person also
has a. Due to the prerequisite relations between the competencies, not all subsets of
competencies are possible competence states (e.g., it is highly unlikely that a person
can multiply but not add numbers). The collection of possible competence states
corresponding to a prerequisite relation is called competence structure (Fig. 1, right
panel). While in the learning/development context we likely find logical prerequisites
(e.g., to add two integers is a logical prerequisite of multiplying two integers), in the
context of using software tools the prerequisite are likely not so obvious (e.g., to
know how to change the font of a text can be considered a prerequisite for knowing
how to make a table). Thus far, the structural model focuses on latent, unobservable
competencies. By utilising mapping functions, the latent competencies are assigned to
a set of tasks/test items/actions relevant for a given domain, which induces a
performance structure, the collection of all possible performance states. Learning or
Fig 1. The left panel shows a prototypical prerequisite function; the right panel
illustrates the corresponding competence structure.
development is not seen as a linear course, equal for all learners; rather, learning
follows one of a set of individual learning paths.
Translated to the context of accomplishing tasks with a software tool, this
approach enables to define the competencies necessary to use the tool in a very
detailed and precise way. By the so-called competence states, the collection of
competencies a specific user has, we can determine which task s/he is able to
accomplish and what kind of information and support is required to solve other tasks.
In addition to the structural hypothesis lying the competence structures, we need an
equally formal idea of problem solving as a process. In this work, we rely on a formal
theory of the human problem solving process, the theory of Newell and Simon ([13];
see also [14] for a conclusive overview).
3.2 A Theory of Human Problem Solving
[13] argued that problem solving is dynamic, rule driven process. The actions a
person performs can be seen as a function of memory operations, control processes,
and rules. The very basis of this approach is to decompose a problem or situation (you
may think about all possible states of the Tower of Hanoi problem) into a problem
space, a collection of all possible and meaningful problem solution states, objects
relevant for a problem, and transition rules, which specify how admissible transitions
from one to another problem solving state can occur. In so far, the approach of Newell
and Simon is highly compatible with the structural approach to learning and
development of CbKST. Consequently, the Tower of Hanoi’s problem space would
include all states where the rules “a larger disk cannot top a smaller” and “all disks
must be on one of the pegs” are not violated (see Fig. 2) Likewise, we establish a
problem space for specific tasks to be performed with a software tool.
3.3 Merging Competence Structures and Problem Spaces
To provide the system with holistic information about the competence domain (i.e.,
the competence structure and the basis of a set of tasks to be accomplished with a
Fig 3. Prototypical problem space for the “Tower of Hanoi” problem.
given tool) and the possible actions within the problem solving processes (i.e., the
problem spaces for the set of possible tasks), both must be linked together.
On a very fine grained level of competencies, with a large number of
possible tasks, and with a large degree of freedom for the user, unfortunately, we
would end up with vast competence structures and problem spaces, which could not
be handled and computed in real time. In the context of games, we introduced the
concept of game states, which we want to term now system state. These are
meaningful accumulations of specific states in the virtual environment (e.g., when
specific objects are assembled to a machine correctly or in terms of regular software,
for example, when a set of options such as font, size or colour are set appropriately
for a specific task). This accumulation is now a substrate of the problem space, which
can be linked to the user’s available and unavailable competencies. There are two
options to do so; first, a deterministic linkage, and second, a probabilistic linkage.
The former means that each system state is associated with a specific set of
competencies that the user has and a specific set of competencies that the user lacks.
The latter means that a numeric value that describes the strength of belief that a
specific set of competencies is available or lacking is assigned to each position
category. Independent of the linkage type, a utility value is assigned to each state to
provide the game with information about “how appropriate” or “how inappropriate” a
system state is for a specific task.
3.4 Updating Competence State Probabilities
In a next step it is necessary to transfer the assumptions of available and lacking
competencies to entire competence states and the likelihood of those states. To
identify a user’s initial (i.e., at first use of the software) competence state, we assume
an initial probability distribution over the competence structure. There exist an almost
arbitrary number of possibilities to obtain such initial probability distribution. The
simplest form would be a uniform distribution in which each competence state has the
same initial probability. The selection of the right method strongly depends on the
given information about the user’s competencies and prior knowledge as well as the
user’s goals.
With each action the user performs, the system updates the probability
distribution over the competence states, where updating rules define the way in which
the probabilities are updated in a specific situation [12]. A simple method is to
increase the probabilities of all competence states that include competencies that are
(either definitely or likely) available when realising the corresponding position
category. Conversely, we can decrease the probabilities of those competence states
that include competencies that are unavailable. On this basis, we are continuously
approaching every action and every realised position category with a clearer
interpretation of the learner’s competence state. Although single interpretations may
not be convincing, with an increasing number of actions, certain and most often
similar competence states become increasingly clear. These procedures are now the
basis for supporting the user in a personalized, suitable, and hopefully in a non
annoying manner.
4 Adaptive Support and Feedback
The basic idea of the micro adaptivity concept is to support the user with suitable
guidance and support and with informative feedback. We have described so far, how
we think that the system can gather a holistic picture of what the user knows, in terms
of which competencies s/he has, and at which point of the problem space the user is
on his/her way to accomplish a specific task (having the idea in mind that
accomplishing a task can be interpreted as solving a problem in a given, well-defined
and rule-based setting). The next logical step is to equip the system with a set of
responses that can be triggered if necessary. Bearing in mind the game-based learning
context, we can distinguish following types of response:
Competence activation interventions may be applied if a user is stuck in
some area of the problem space and some competencies are not used even
though the system assumes that the user possesses them. For example, the
system can make a specific function (e.g., a tool bar button) more visible.
Competence acquisition interventions may be applied in situations when the
system concludes that the user lacks certain competencies. For example, the
system informs the learner about specific features of the software and how to
use them.
Feedback may be utilised to provide the user with information about the
work progress. For example, the software could inform the user if s/he has
used the software in an efficient way.
Assessment clarification interventions may be applied, for example, in the
form of a query if the user’s actions provide contradicting support for the
assumption of a certain competence state. In addition, such type of
intervention maybe used to determine which task a user wants to accomplish.
Such kind of support and guidance proved to be effective in the learning context
[15], more importantly in the context, recent research showed that such interventions
could improve immediate goal achievement [16].
5 Leering at the Games: 80Days
5.1 Architecture
The concept of micro adaptive assessment and interventions has been developed in
the context of the ELEKTRA project and its successor 80Days. In 80Days, prototype
game was developed to demonstrate and evaluate the concept of micro adaptivity. The
architecture consists of four modules, (1) the game itself, quite traditionally, is created
using a high-end gaming engine, (2) a skill assessment engine updates the competence
state probabilities, and the resulting information regarding the learner’s competence
state and its changes are then forwarded to (3) an educational reasoner, the
pedagogical part of micro adaptivity. Based on a set of pedagogical rules and meta-
rules as well as learning objectives, this engine provides recommendations on
adaptive interventions to the adaptation realisation module. This, in turn, maps the
abstractly formulated educational recommendations onto more concrete game
recommendations. In this mapping process, data on game elements and information
on previously given recommendations are considered. The necessary information for
the assessment-intervention loop is stored (4) in an OWL [17], which allows the
aforementioned engines to extract not only information, but also the relationships
among the information from the ontology.
5.2 The Game
We have developed a learning adventure that is supposed to teach geography for the
age group of 12 to 14 years. The curriculum includes, for example, knowledge about
the planet Earth such countries or cities but also aspects such as longitude or latitude.
In the game the learner takes the role of a boy or a girl (depending on the learners’
gender) at the age of 14. The story starts from an extraordinary event; a space ship is
landing in the backyard and an alien named Feon appears. Feon turns out to be a
friendly alien, being an alien scout who has to collect information about foreign
planets, in this case planet Earth. The learner accompanies Feon and is having fun
with flying a space ship and exploring interesting places on Earth. Feon creates a
report about the Earth and its geographical features. This is accomplished by the
player by means of flying to different destinations on Earth, exploring them, and
collecting and acquiring geographical knowledge. The goal is to send the Earth report
as a sort of travelogue about Earth to Feon’s mother ship. At a certain point in the
game, however, the player makes a horrible discovery; the aliens are not really
friendly but collect all the information about Earth to conquer the planet, lately. This
discovery reveals the “real” goal of the game: The player has to save the planet and
the only way to do it is to draw the right conclusion from the traitorous Earth report.
The subject matter is embedded in hat story and learning occurs at various events in
the game. From a pedagogical point of view, learning occurs by receiving information
(e.g., seeing/reading something in the game or hearing something from Feon or other
game characters), problem solving (e.g., reducing the negative impacts of a flood by
appropriate “terra-forming”), or imitation (e.g., watching other game characters and
learning from their behaviours). Screenshots are shown in Fig. 3.
Fig 3. Screenshots from 80Days’ game demonstrator.
6 Conclusion and Outlook
The idea of adaptation on the micro level is to monitor a learner’s behaviour within an
adaptive DEG and to provide him/her with appropriate and tailored support and
feedback. In the context of the 80Days project we implement the corresponding
functionalities in a compelling demonstrator game. This game was subject of
experimental evaluations which showed that the envisaged kind of “subliminal” or
“hidden” assessment not only works in terms of assessment accuracy but also that the
corresponding interventions lead to superior learning performance, motivation, and
satisfaction with the game in comparison to no adaptive interventions or inappropriate
interventions [18].
In the present paper we raised the idea that such non-invasive assessment method
and especially the underlying logic might be successfully applied ion the context of
regular software as well. While the famous and unsuccessful even annoying paper clip
assistant of Microsoft’s Office was a good idea in its essence, the underlying logic
was not smart enough to convince most of its users. If we can transfer the
psychologically sound frameworks of CBKST in combination with problem spaces
we can equip a software system with a “deeper understanding of the user and his/her
needs – at least in comparison to other support methods.
Of course, this approach is still in its infancy and must be elaborated in future
work. Exactly here lies the major purpose of this paper, that is, introducing the
concept to a community broader than that of game-based learning, to encourage
discussions and to broader the concept for a variety of applications.
Acknowledgments
The research and development introduced in this work is funded by the European
Commission under the seventh framework programme in the ICT research priority,
contract number 215918 (80Days, www.eightydays.eu).
References
1. de Freitas, S.: Learning in immersive worlds. A review of game-based learning (2006).
Retrieved August 28, 2007 from http://www.jisc.ac.uk/media/documents/programmes/
elearning_innovation/gaming%20report_v3.3.pdf
2. Prensky, M.: Digital game-based learning. New York: McGraw-Hill (2006)
3. Van Eck, R.: Digital game-based learning. It’s not just the digital natives who are restless.
Educause Review, March/April 2006
4. Kickmeier-Rust, M. D., Peirce, N., Conlan, O., Schwarz, D., Verpoorten, D., & Albert, D.:
Immersive digital games: Interfaces for next-generation e-learning. In C. Stephanidis (Ed.),
Universal Access in HCI, Part III, HCII 2007, LNCS 4556 (pp. 647–656). Berlin: Springer
(2007)
5. Kafai, Y. B.: Playing and making games for learning. Games and Culture, 1(1), 36-40
(2006)
6. Csikszentmihalyi, M.: Flow: The psychology of optimal experience. New York: Harper and
Row (1990)
7. De Bra, P. M. E.: Adaptive hypermedia, In H. H. Adelsberger, Kinshuk, J. M. Pawlowski, &
D. Sampson, (Eds.), Handbook on Information Technologies for Education and Training
(pp. 29-46), Berlin: Springer (2008)
8. Kickmeier-Rust, M. D., Albert, D., Hockemeyer, C., & Augustin, T.: Not breaking the
narrative: Individualized Competence Assessment in Educational Games. In Proceedings of
the European Conference on Games-based Learning (ECGBL), October 25-26, 2007,
Paisley, Scotland (2007)
9. Albert, D., & Lukas, J. (Eds.): Knowledge spaces: Theories, empirical research, and
applications. Mahwah, NJ: Lawrence Erlbaum Associates (1999)
10. Korossy, K.: Extending the theory of knowledge spaces: A competence-performance-
approach. Zeitschrift für Psychologie, 205, 53-82 (1997)
11. Doignon, J.-P., & Falmagne, J.-C.: Spaces for the assessment of knowledge. International
Journal of Man-Machine Studies, 23, 175-196 (1985)
12. Doignon, J.-P., & Falmagne, J-C.: Knowledge spaces. Berlin: Springer (1999)
13. Newell, A. & Simon, H.: Human Problem Solving. Englewood Cliffs, NJ: Prentice-Hall
(1972)
14.Newell, A.: Unified Theories of Cognition. Cambridge, MA: Harvard University Press
(1990)
15. Azevedo, R., & Bernard, R.M.: A meta-analysis of the effects of feedback on computer-
based instruction. Journal of Educational Computing Research, 13, 111-127 (1995)
16. Tan, J., Biswas, G., & Schwartz, D.: Feedback for metacognitive support in learning by
teaching environments. In Proceedings of the 28th Annual Meeting of the Cognitive Science
Society, Vancouver, Canada, pp. 828-833 (2006)
17. Kickmeier-Rust, M. D., & Albert, D.: The ELEKTRA ontology model: A learner-centered
approach to resource description. Advances in Web Based Learning – ICWL 2007 (pp. 78-
89). Lecture Notes in Computer Science, 4823/2008. Berlin: Springer (2008).
18. Kickmeier-Rust, M. D., Marte, B., Linek, S. B., Lalonde, T., & Albert, D.: The effects of
individualized feedback in digital educational games. In T. Conolly & M. Stansfield (Eds.),
Proceedings of the 2nd European Conference on Games Based Learning (pp. 227-236),
October 16-17, 2008, Barcelona, Spain. Reading, UK: Academic Publishing Limited (2008)
Article
Digital learning games are designed to foster both student learning and enjoyment. Given this goal, an interesting research topic is whether game mechanics that promote learning and those that promote enjoyment have different effects on students’ experience and learning performance. We explored these questions in Decimal Point, a digital learning game that teaches decimal numbers and operations to 5th and 6th graders, through a classroom study with 159 students and two versions of the game. One version encouraged playing and learning through an open learner model (OLM, N = 55), while one encouraged playing for enjoyment through an analogous open enjoyment model (OEM, N = 54). We compared these versions to a control version that is neutral with respect to learning and enjoyment (N = 50). While students learned in all three conditions, our results indicated no significant condition differences in learning outcomes, enjoyment, or engagement. However, the learning-oriented group engaged more in re-practicing, while the enjoyment-oriented group demonstrated more exploration of different mini-games. Further analyses of students’ interactions with the open learner and enjoyment models revealed that students who followed the learner model demonstrated better in-game learning and test performance, while following the enjoyment model did not impact learning outcomes. These findings indicate that emphasizing learning or enjoyment can lead to distinctive game play behaviors, and that open learner models can be helpful in a learning game context. In turn, our analyses have led to preliminary ideas about how to use AI to provide recommendations that are more aligned with students’ dynamic learning and enjoyment states and preferences.
Article
In the past decade, a number of personalized learning systems have been developed and they mainly focus on learners' prior knowledge. On the other hand, previous research suggested that gender differences and cognitive styles have great effects on student learning. To this end, this study examines how human factors, especially gender differences and cognitive styles, affect learners' reactions to a personalized and non-personalized learning systems based on learners' prior knowledge. Forty-four university students participated in this study. The results show that females and Serialists show positive reactions to the personalized learning system, while males and Holists demonstrate similar reactions to the personalized learning system and the non-personalized learning system. The implications of these results for the design of personalized learning systems are discussed.
Article
Full-text available
A quantitative research synthesis (meta-analysis) was conducted on the literature concerning the effects of feedback on learning from computer-based instruction (CBI). Despite the widespread acceptance of feedback in computerized instruction, empirical support for particular types of feedback information has been inconsistent and contradictory. Effect size calculations from twenty-two studies involving the administration of immediate achievement posttests resulted in a weighted mean effect size of .80. Also, a mean weighted effect size of .35 was obtained from nine studies involving delayed posttest administration. Feedback effects on learning and retention were found to vary with CBI typology, format of unit content and access to supplemental materials. Results indicate that the diagnostic and prescriptive management strategies of computer-based adaptive instructional systems provide the most effective feedback. The implementation of effective feedback in computerized instruction involves the computer's ability to verify the correctness of the learner's answer and the underlying causes of error.
Article
Full-text available
Note: This article was UPDATED and revised in 2015 in a new article entitled "DGBL: Still Restless After All These Years" which can be found in Research Gate and at Educause Review. What follows are BOTH abstracts: 2006 Abstract: After years of research and proselytizing, the proponents of digital game-based learning (DGBL) have been caught unaware. Like the person who is still yelling after the sudden cessation of loud music at a party, DGBL proponents have been shouting to be heard above the prejudice against games. But now, unexpectedly, we have everyone’s attention. The combined weight of three factors has resulted in widespread public interest in games as learning tools. 2015 Abstract: Nearly a decade ago, I wrote an article for EDUCAUSE Review about digital game-based learning (DGBL) and the challenges it faced.1 I suggested that once proponents of DGBL were successful in convincing people that games could play a role in education, they would be unprepared to provide practical guidance for implementing DGBL. Just as when the person shouting to be heard at a party is suddenly the center of attention at the moment there is a lull in the conversation, we DGBL proponents had everyone's attention—but not much to say. In the article I also suggested that our sometimes overzealous defense of videogames (hereafter often referred to as "digital games") ran the risk of overselling the benefits (and underreporting the challenges) of using digital games in formal education. Digital games, I said then and still believe today, are effective as embodiments of effective learning theories that can promote higher-order outcomes. Our inability to provide guidance in doing so a decade ago was ceding the DGBL front to digital games as tools for making didactic, instructivist learning (i.e., lectures) more "engaging." DGBL, I suggested, was effective not as a means for making learning "fun" or for "tricking" students into learning; DGBL was effective because it supported powerful learning strategies such as situated learning, authentic environments, and optimized challenge and support (scaffolding). What was needed was a renewed focus on (1) research about why DGBL is effective and (2) guidance on how, when, for whom, and under what conditions to integrate digital games into formal education. I was not the only one with these ideas, but my timing and the venue combined to reach many people. That 2006 article has been cited more than 1,000 times since then.2 Yet though these ideas continue to resonate with many people, much has changed in terms of research, practice, and to some extent, my own beliefs about the future of DGBL.
Article
Full-text available
This article presents an overview of what we know about two perspectives, coined instructionist and constructionist, to games for learning. The instructionists, accustomed to thinking in terms of making instructional educational materials, turn naturally to the concept of designing instructional games. Far fewer people have sought to turn the tables: by making games for learning instead of playing games for learning. Rather than embedding “lessons” directly in games, constructionists have focused their efforts on providing students with greater opportunities to construct their own games—and to construct new relationships with knowledge in the process. Research has only begun to build a body of experience that willmake us believe in the value of playing and making games for learning.
Article
Full-text available
Past research on feedback in computer-based learning envi- ronments has shown that corrective feedback helps immediate learning, whereas guided and metacognitive feedback help in gaining deep understanding and developing the ability to transfer knowledge. Feedback becomes important in discov- ery learning environments, where novice students are often overwhelmed by the cognitive load associated with learning and organizing new knowledge while monitoring their own learning progress. We focus on feedback mechanisms in teachable agent systems to help improve students' abilities to monitor their agent's knowledge, and, in the process their own learning and understanding. Our studies demonstrate the effectiveness of guided metacognitive feedback in preparing students for future learning.
Article
The book presents the case that cognitive science should turn its attention to developing theories of human cognition that cover the full range of human perceptual, cognitive, and action phenomena. Cognitive science has now produced a massive number of high-quality regularities with many microtheories that reveal important mechanisms. The need for integration is pressing and will continue to increase. Equally important, cognitive science now has the theoretical concepts and tools to support serious attempts at unified theories. The argument is made entirely by presenting an exemplar unified theory of cognition both to show what a real unified theory would be like and to provide convincing evidence that such theories are feasible. The exemplar is SOAR, a cognitive architecture, which is realized as a software system. After a detailed discussion of the architecture and its properties, with its relation to the constraints on cognition in the real world and to existing ideas in cognitive science, SOAR is used as theory for a wide range of cognitive phenomena: immediate responses (stimulus-response compatibility and the Sternberg phenomena); discrete motor skills (transcription typing); memory and learning (episodic memory and the acquisition of skill through practice); problem solving (cryptarithmetic puzzles and syllogistic reasoning); language (sentence verification and taking instructions); and development (transitions in the balance beam task). The treatments vary in depth and adequacy, but they clearly reveal a single, highly specific, operational theory that works over the entire range of human cognition, SOAR is presented as an exemplar unified theory, not as the sole candidate. Cognitive science is not ready yet for a single theory – there must be multiple attempts. But cognitive science must begin to work toward such unified theories.