Label Cloud

Sunday, April 24, 2011

Who plays, how much, and why?

Williams, D., Yee, N., & Caplan, S. (2008). Who plays, how much, and why? Debunking the stereotypical gamer profile. Journal of Computer-Mediated Communication, 13, 993-1018. http://onlinelibrary.wiley.com/doi/10.1111/j.1083-6101.2008.00428.x/full

The paper is about a study designed to find the demographics of MMOs players. The research focuses on age, gender, education, other media use, religion, income, and race and compares these with general population.

Bartle's player taxonomy: Achievers, Socializers, Explorers (enjoy discovering the geographical boundaries of the world, analyzing rules), and Killers
Yee's set of motivation items for play is added to Bartle's:
Achievement, Relationship, Manipulation, Immersion, Escapism
-three second order factor structures:
1-Achievement: 
  • advancement, 
  • analyzing game mechanics,
  • competition
2-Social: 
  • chatting and casual interactions
  • developing supportive relationships
  • teamwork
3-Immersion:
  • geographical exploration
  • role-playing
  • avatar customization
  • escapism
Sherry's topology of motivations for video gamers : has Arousal as addition to above motivations

Research questions:
-What is the distributions of age, gender, race and class among MMO players?
-How do the demographics of MMO players compare to general population?
-Who plays how much?
-What are the media use pattern of MMO players?
-How does the physical health of MMO players?
-How does the mental health of MMO players?
-What are the motivations of MMO players?
Method:
Physical health measured by three indicators: BMI, exercise habits and physical impairment
Mental health measured by asking whether they are diagnosed by depression, substance addiction, or anxiety 
Results:
age: mid 30s, counter to stereotypes (not teens)
gender: 80% male, female players play more hours per week 
race: White, Native Americans
religion: different levels of spirituality but mostly players have no religion
media use: watching TV vs playing online, players watch less TV then general population
health: players are physically healthier but mentally less healthier
motivation: players were motivated to play for achievement, immersion and social reasons- achievement was stronger
-adult, male, in his 30s, white Native American, middle class

why highly achievement oriented? 
-person may lack that achievement in real life
-to achieve high self-esteem
why socially oriented?
-when threatened, something upset in real life

Next steps for researchers is:
- to develop and extend theories that fit these findings 
- to question why stereotypes about gamers formed



Wednesday, April 20, 2011

SodaConstructing Knowledge

Svarovsky, G. N. & Shaffer, D. W. (2007). SodaConstructing Knowledge through Exploratoids. Journal of Research in Science Teaching, 44, 133-153.

The paper describes a preliminary study about engineering design activities for K-12 students and the work on microworlds as learning tools. It focuses on design- build- test (DBT) cycle.
Based on Petroski’s argument that states children are born engineers, and the pedagogical praxis theory that suggests authentic recreations of professional practices for designing technology-based learning environments,  paper suggests that to understand the key concept in physics, children should design as engineers, not with physical materials, but with computer simulations, which are cheaper and faster to design, and safer. Since it is less time consuming, it increases the iterativity of the DBT cycles.

The experiment is done with SodaConstructor, a spring-mass modeling simulation.
Engineering design is of 3 stages
1) Conceptual, - brainstorming
2) Preliminary - alternatives are modeled and analyzed
3) Detailed - the selected alternative is gone through final design
The DBT cycle is an iterative process of:
1) develop and evaluate design alternatives
2) design a solution to the problem
3) build a prototype, test it
Learn by Design (LBD): iteratively messing around, whiteboarding

Microworlds
Simulations are of a computational microworld
Two key factors:
1)autoexpressivity: An autoexpressive microworld gives different responses (feedback) to a student’s actions depending on the extent to which the student is explicit about his or her intentions
2)expressivity, an affordance users experience when interacting with the tool. When the students explore the microworld, it is both meaningful and motivating for them , affording them a sense of a control.

Experiment:
2 workshops, 6 students from sixth grade, and 6 seventh grade, total of 10 male, 2 female, heterogeneous group .
SodaConstructor microworld: spring mass modeling system where students can design structures and test them against gravity.
Followed LBD curriculum:
- worked individually to design
- discuss their design within their team
- redefine & redesign
- choose best design
- whole group discussion
Data collection: Pre & Post-Interviews, Videotape during the workshop
Definition of center of mass were asked in both pre and post interviews, and the understanding of center of mass decreased after the workshop.
Students feedback about the SodaConstructor:
One student iterated through 19 DBT cycles to find the right design. The iterative process of rapid prototyping in an autoexpressive microworldmade it possible for him to incrementally develop scientific understanding.

Design factors for educationally effective animations and simulations

Plass, J.L., Homer, B., & Hayward, E. (2009). Design Factors for Educationally Effective Animations and Simulations. Journal of Computing in Higher Education, 21(1), 31-61.

Dynamic visuals vs. Static images:
Dynamic visual environments’ educational effectiveness depends on multitude of design considerations involved in development of effective visual materials for learning.
According to Hoffler and Leutner’s studies, dynamic visualizations are more effective than static images when dynamic visualizations are representational rather than decorative, and also when the target knowledge is procedural motor knowledge rather than declarative knowledge.

Visual perception and cognitive processing of visual information:
Because only a small amount of visual information is available to the retina can be processed, objects compete for representation and processing. The outcome of this competition determines which objects are perceived and defines the visual attention.

Dual coding theory: compares concurrent processing of info in both verbal (system for verbal info) and non-verbal (system for images) systems with processing it in one system only.

Cognitive load theory: how processing of information and constructing knowledge works in a limited working memory resources. Three types of load: Intrinsic(inherent complexity of info), extraneous(unnecessary processing of unrelated info), germane(mental effort to understand)

Cognitive theory of multimedia learning: suggests that learners first select relevant info and then organize it into coherent verbal and visual mental representation, then integrate these with one another, and with prior knowledge.

Integrated model of text and picture comprehension: cognitive processing relies on multiple memory systems with limited capacity. Schnotz distinguishes processing of symbolic info and iconic info, which results in the mental models: depictive (visualozation, iconic) and descriptive(symbolic, textual).

Multimedia principle: Comprehension and transfer are enhanced when text is accompanied by pictures compared when text alone.
Modality principle: When text and visualizations are presented together, learners experience higher cognitive load, and less understanding of the material when text is visually presented compared to as narrative.

Design principles for effective dynamic visualization

Visual design principles:
Split-attention principle: where learners split their attention in between e.g. movie with subtitles.
To avoid this effect, designers can integrate the sources and arrange the timing of their presentations.

Contiguity principle:  how presented related information close to one another, enhances learning by reducing visual search tasks.
Spatial contiguity principle: spatial arrangement of the information
Temporal contiguity principle: timing of the presentations

Cueing principle (Signaling): addition of design elements direct the learner’s attention to the important aspect of the learning material.

Representation type of information:
An emerging key principle for designing dynamic visualizations is representing in iconic form rather then only in symbolic form.

According to the Cognitive Load Theory processing depictive info requires less mental effort than descriptive because depictive info by definition relates directly to their referent, but descriptive info needs to be interpreted.

Written vs pictorial instructions for building molecular models: for simple material  molecular models, both were effective in the same amount. But for complex molecules, pictorial was more effective because the pictorial representation reduces cognitive load ,freeing cognitive resources and allowing student to solve complex tasks.

Simulation of Kinetic Theory of Heat, Symbolic vs Iconic: under high cognitive load, iconic representation improved the overall understanding

Color Coding: using color to highlight important features if the visual displays and to draw connections between multiple sources of information hence resulting in reductions in working memory and search demands. e.g colorful maps.
Color coding can eliminate the split-attention effect.

Integration of multiple dynamic visual representation:  Integrated and dynamically linked are the most effective among the other representations: 1) separate and dynamically linked, 2) separate and non-linked.

Interaction design principles:
Three levels of interactivity, 1) control of the delivery (click on a button),  2) manipulation of the content (setting parameters), 3) control of the representation (rotating an object)

Research about interactive simulations show that control over the representation of the info help learner comprehend better.

Learner control-segmenting principle: describes how learners’ comprehension of material is better when they can control the segmented presentation rather than continuous presentation.

Guided-discovery principle: Guidance decreases the extraneous cognitive load demands on the learnerby supporting the learners’ abilities to organize and integrate new info. Feedback is another form of guidance discovery based learning.

Learner control-pacing: learning is improved when learners are given control over the pacing of information.(start, pause, stop).

Task appropriateness: Efficiacy of a simulation depends on the degree on which it is in line with learning objectives, so they can prepare the learner for future tasks. Visualizations should have a interpretational nature.

Manipulation of  content: Learning from visualizations is improved when learners have the ability to manipulate the content. According to the ideal gas law experiment,  manipulation level interactivity increased learners’ germane cognitive load compared to  providing only control over pacing of the materials.

Future research :
-needs to ask questions for specific dynamic visualizations, instead of looking at animations, and simulations in general.
-needs more systematic approach rather than current theoretical approach.
-interaction design needs a detailed typology  of levels of interactivity
-should investigate the cognitive load related questions of visual aspects of the model progression.
-should focus on information design:
How to improve information design:  we need to better integrating results from visual perception and visual cognition. We need to link neuroscience, visual attention, cognitive load and learning, visual representations and emotions in learning.

Conclusion:
The research in this paper have to be connected with work in the area of neuroscience and cognition.

Evaluating and Managing Cognitive Load in Games

Kalyuga, S., & Plass, J.L. (2008). Evaluating and managing cognitive load in educational games. In R.E. Ferdig (Ed.), Handbook of Research on Effective Electronic Gaming in Education, 719-737. New York: IGI Global.
The paper talks about our cognitive architecture and its implications for games for learning.
Processing limitations of working memory represent a major effect in learning and performance. Lots of search to make connections between sources of info, irrelevant info, etc due to poor design interfaces makes use of cognitive resources and causing high cognitive load.
Levels of learner prior knowledge and experience in domain is a major factor while designing for learning and considering cognitive load.
Software applications including educational games need to have understandable and recognizable and functionally efficient interface components.

Our cognitive architecture: LTM (Long-term memory), and WM (working memory)
1) LTM is the large store of organized info with unlimited storage capacity and duration
2) WM is the functional mechanism that limits the scope of immediate changes to the LTM. It is limited in duration and storage.
3) During the learning process, information is borrowed from other sources by actively being reconstructed in WM, reorganized and integrated with available knowledge in LTM.
4) Ability to organize complex situations and tasks.
cognitive economy principle: tends to minimize cognitive resources involved in performance of a cognitive task.

To reduce cognitive load in learning
- use of available knowledge as a guide and external function
- Encapsulating information elements into chunks in WM using the prior knowledge structures in LTM e.g. phone number by subsets such as one’s birth year...
- practicing skills, until they become automated
- learner’s expertise change the reduction in cognitive load
-  direct instructions and guidance (Plass et al, 2007) found that simulation exploration outperforms the direct instructions.

Intrinsic cognitive load: caused by cognitive activities that establish key connections between elements of information and integrating them with available knowledge structure and building new knowledge structures in WM. This type of cognitive load is depends on the degree of interactivity and learner’s level of expertise.

Extraneous load: diversion of cognitive resources on activities irrelevant to performance and learning.Caused by design related factors. Searching for solutions, game rules, evaluating game states... Extraneous load imposed by:
1) representations that require users extensive search
2) excessive rate of information change that introduces too many new elements to WM
3) insufficient external guidance causing search
4) user knowledge overlaps with provided external guidance, redundancy

For educational games,
- sufficient instructional guidance and support for learners is important
- benefits of guidance,
- students’ unfamiliarity with the game hardware
- no instruction leads to discovery based learning, but guided games is more instructionaly effective than discovery based games
- physical simulations benefit from addition of iconic representations for low-prior knowledge learners.
- interactive manipulations have benefit for experienced learners but cause extraneous cognitive load.

How to compare cognitive load?
- asking users how difficult
- dual - task technique, secondary test could be reaction times to mouse clicks, counting backwards
- think aloud protocols

Design to Learn about Complex Systems

Hmelo, C.E., Holton, D.L., & Kolodner, J.L. (2000). Designing to learn about complex systems. Journal of the Learning Sciences, 9(3), 247-298

The paper is about how to enhance the learning of the complex systems, such as car engine, human respiratory system, etc. Complex systems are the ones that are not actually being fully understood by the children. The paper suggests design activities in order to help children acquire a deeper understanding of such complex systems. In the paper, an experiment is reported. In the design experiment, 6th grade children are asked to design artificial lungs in order to let them learn human respiration in a better way.
The paper builds on Perkins’ “knowledge as design“approach, in which Perkins suggested helping students view systems as designs instead of defining the parts, and memorizing the definitions as in the traditional approach. Perkins’ approach addresses the functional roles; the mechanisms by which the roles are carried out; and the way the functions are interact with each other. However it doesn’t encourage the students to be engaged in such learning. Learning by Design, LBD, approach helps the students to be involved actively with the design and modeling activities to learn. Goals of the LBD approach:
1)      The extent which a design approach could be used to help children learn.
2)      Examine Structure-Behavior-Function(SBF) relations
3)      Investigate ways of implementing a design approach, answering the questions: What design challenges? How help students remain focused? How to endure the content is covered? How to organize the activity to promote deep understanding?
Why are complex systems difficult to be covered?
Systems are dynamic entities, their organizational levels are difficult to visualize. Also, one reason could be the way the individuals are introduced to these systems in the traditional way.
Structure-Behavior-Function (SBF):
Structure refers to the physical structure, e.g. lungs in a respiratory system
Function refers to the purpose of the system. E.g. respiratory system transports oxygen throughout the body to the organs that require it.
Behavior refers to the dynamic mechanism and workings that allow the structures to carry out their function. Problem of understanding the behavior of a system is because it is most of the time invisible (e.g electrical impulses traveling through nerves) and it is time-delayed causality.
So through design challenges and modeling activities that focus on behavior and function rather than just structure would help the students move form novice to more expert understanding of systems.
Designing affords such understanding
(Lehrer, Perkins)By designing students are constructing rather than receiving knowledge.
(Kolodner et al)During the process of designing, the students construct, apply, and evaluate models. They test and discuss their design with each other.
Linn suggests the need for multiple iterations during modeling to help them construct progressively. In this way, they can seek answers to their problems during the modeling. They test and seek out better models.
Implementing LBD
When children conduct experiments they usually care about creating the outcome rather than the understanding, since the outcome is a concrete product whereas understanding is more abstract.
To be able to promote LBD, one has to:
1)      Find a balance between having students work on design activities and reflecting.
2)      Give real world knowledge without overwhelming the students with irrelevant aspects of the world.
3)      Maintain the understanding the concept more important than completing the task.
Authors’ initial approach was problem-based learning (PBL), an implementation that focuses on learning by complex problem solving activity.
How does PBL session work?
FACTS: While working on understanding the problem students stop to reflect on the data they collect
IDEAS: potential solutions to the problem they are working on
LEARNING ISSUES: concepts they need to learn more about to solve the problem
ACTION PLAN: they develop a plan for proceeding
Since PBL was not enough to manage construction and testing activities, authors added case-based reasoning (CBR) to the PBL implementation, to keep students focus on the design challenge. CBR focuses on storing problem solving experiences (cases), indexing them so they could be found by ways of searching and adapting the solutions to old cases to solve new problems. This shows that learning is an iterative process: for deep understanding one has to fail, and then need to explain why failed, because he is motivated to explain the failure once he fails.
CBR also keeps pointers such that one can transfer what he has learned in a situation to a new situation.
In the class, students need to actually build and test based on their understanding, and get feedback from testing, then explain what happened, and revise the understanding and try again. This leads to multiple cycles of designing, constructing, testing, explaining, and revising to the PBL framework.

Authors’ original conception of LBD (by using PBL together with CBR):
Groups of 3 or 4, for construction, testing, redesign activities;
Planning, monitoring done as a whole class activity facilitated by the teacher;
A whiteboard as the device for recording shared experiences, ideas, sights, questions.

The Design Experiment
There were 42 students, 6th grade, heterogeneous group, who spent 2 weeks for the experiment, in two classrooms, one is LBD setup and other is comparison classroom.

The students were asked to solve a design problem, which asked them to design an artificial lung and build a model of some piece of the design.

The kids are given resources online/books about human respiratory system. For the PBL process, the teacher wrote Fact, Idea, Learning Issue, and Active Plan on the whiteboard. They reviewed whiteboard at the beginning of each class period, and updated the PBL columns as they discussed. They ended each day by reviewing, updating, or recreating.

Hands-on Investigation: they used the devices used in hospitals and examined, who had the greatest lung, whether size or different activities change the air moving in and out. Then the teacher wanted them to make connections between their design challenge and the hands on investigation.
On day 6, they began designing. There were materials such as balloons, clay, etc. Some experimented with these materials, some copied designs from books.

Analysis: Intended vs. enacted curriculum
Students did important iterations that improved their models to come to a good understanding of the concept. However, they stopped when each had a partial working model. They had discussions after that point, but they didn’t apply the new ideas to their models.
Design and modeling: They were motivated by the design challenge but since it was very complex they couldn’t apply what they learned. Problem here was the poor connection between their research, which focused on the big issues and modeling, which needed more details.
Since they couldn’t have lots of iterations of their models, they didn’t discuss behavior and function.
What students learned:
Post-tests were done to compare with the control group. Pre and post test results couldn’t inform how the mental model of the kids had changed, but the results showed that LBD students had a better understanding of the lungs and respiratory system than the control group.

Affordances of Design:
Lung design challenge was able to: Generate questions, coming up with ideas, making connections between science and its usefulness in the world
Not able to: Coming up with solutions to be tested, focus and guide an investigation
If the authors redo this experiment:
They would choose the right construction materials
Include a discussion of what a model is, what we can use them for
Motivate students move iteratively design of better and better models


Even though it was a long paper, the design experiment itself was the long part, but that is actually fun to read. I like the approach the authors used for the experiment which is Learning Based Design (LBD), that is a mix of Problem Based Learning (PBL) and Case Based Reasoning(CBR).

Cognitive Design Factors

Egenfeldt-Nielsen. Thoughts on learning in games and designing educational computer games.

1- Using educational games for formal learning at school
2- Using entertainment games to motivate learning at school
3- Using educational games for informal learning during leisure time
Games constructed for learning : edutainment: not as fun as commercial games
So designers , researchers and teachers should come together to bring learning aspects to commercial games, the games that kids love to play!
Game should be cool! Learning in a game should not make us forget the game part, which is fun.
-Fun vs Learning? How to balance these two?
Learning by Bateson (Bateson, 1972: 283)
process going from a little change to larger change influencing the behavior of the individual.
Zero Learning -> Learning I -> Learning II -> Learning III
Zero Learning: the simple receipt of information, In a game, a mouse move activates something.
Learning I: trial- error. In a game, finding out using a unit against another unit is a bad idea, try another available unit.
Learning II: kind of meta-learning. In a game, changing your playing style from player-killing to role-playing.
Learning III: meta-meta-learning. Rare.
Learning real life vs. Learning conceptually
Learning real life elements are facts, behaviors, communication, skills ,theories, and language.
Learning conceptually concepts are reasoning, process, procedures, creativity, system understanding, If talking about learning in games, talking about learning conceptually. Because, in this one, you reflect in a situation through analysis, and reasoning.
Counter Strike: move and shoot. Learning process of the game is not different than daily activities, but a curriculum is missing.
Europa Universalis: covers European History primarily. Philosophy of the game is to make differences in history.
Layered vs. Branched Approach
Layered presents options to player to choose a new area of the game. You have everything , all the time. Strategic and simulation games.
Branched gives different routes to be chosen at specific times. Expands slowly. Role-playing games.
Three basis factors: Play, knowledge and story:
Learning occurs when knowledge becomes story and story becomes knowledge.
-Games should be constructed to facilitate learning not to force learning

Interactivity in Multimedia Learning

Domagk, S., Schwartz, R., & Plass, J.L. (in press). Interactivity in Multimedia Learning: An integrated Model. Computers in Human Behavior. 

Constructivist approach for learning: learner should be actively engaged in the process of knowledge construction
-interactive multimedia learning environment lets learner be actively involved in the process


What is interactivity?
Interactivity in the context of computer-based multimedia learning is reciprocal activity between a learner and a multimedia learning system, in which the reaction of the learner is dependent upon the reaction of the system and vice versa.

Moreno and Mayer's five types of interactivity: 
1-dialoguing: learner receives questions and answers or feedback
2-controlling: control over pace
3-manipulating: control over aspects of the presentation
4-searching: entering queries, selecting options
5-navigation: selecting information sources 

Kalyuga's three types of learner control afforded by the learning system:
1-control over information delivery, 
2-representational forms
3-content
 and two dimensions: flexibility and dependence

Different suggestions on interactivity: 
- from non-interactive to highly interactive
- interactivity may cause large extraneous load because of the extra information need to be process or split attention that interferes with learning

Kennedy introduces cognitive interaction model
-continuous feedback between behavioral processes, instructional events and cognitive processes
-Problem: doesn't consider learner's characteristics, emotions
-Solution: INTERACT: Integrated Model of Multimedia Interactivity
Six principles:
1-learning environment: includes the instructional design and the affordances of the learning system
2-behavioral activities: what the learner does physically to interact with the system, 
3-cognitive and metacognitive activities
       -cognitive: mental operations, procedures and processes which learner perform to select, mentally integrate, organize and integrate new information into coherent knowledge structure.
  • surface-level processing: use of cognitive strategies of repetitive rehearsal and rote        memorization to encode new information into working memory- learning outcomes: retention and recall.
  • deep-level processing: results from use of cognitive strategies as elaboration, self regulation, critical thinking to integrate new information with prior knowledge - deep understanding and transfer
  • cognitive process dimensions: (Bloom's taxonomy)-remember, understand, apply, analyze, evaluate and create
  • knowledge dimensions: factual, conceptual, procedural and metacognitive knowledge
      -metacognitive: knowledge about cognition and regulation of cognitive activities

4-motivation and emotion: conditions of the learner that arise from given situation

  • motivational factors can affect cognitive engagement
  • metacognitive factors can regulate cognitive processing and affect 
  • in INTERACT, emotion, motivation and cognition are intertwined and they are in a feedback loop, influencing and being influenced by cognition
5-learner variables: prior knowledge, learner characteristics- influence level of engagement, learner's motivation, etc
6-learner's mental model: exiting knowledge structures that learning brings to learning activity and the knowledge that learner gains as a result of the learning activity

Educational design features in INTERACT:
Learner control: can help the learner to adjust information to his cognitive needs
Guidance/feedback: guidance: directs cognitive processes of the learner, generating hypothesis, monitoring, and structuring
  • interpretative: guides the learner in structuring knowledge by activating the prior knowledge
  • experimental: helps the learner to set up  and interpret experiments
  • reflective: helps  learners in reflecting on the learning process and new information
      feedback: provide different kinds of instructional support