NSF Awards: 1542970
The MIT Imagination, Computation, and Expression Laboratory sees access to high quality STEM education as a social justice issue of our time. Through this project, we seek to discover best practices for using virtual identities (from avatars to social media profiles) to enhance performance, engagement, and STEM identity development for diverse secondary school computer science (CS) students.
We designed and deployed workshops in public schools to catalyze excitement and learning about CS, foster critical thinking skills pertinent to information technologies (e.g., surveillance, privacy, big data, etc.), and support development of students’ identities as empowered computational learners and practitioners. Our vision reflects an anti-deficit ideological stance towards students from underrepresented groups, seeing their social identities not as obstacles, but as positive resources for their STEM identities.
Our workshops use our custom platform called MazeStar that allows students to create customized games addressing issue within their communities. MazeStar contains a CS learning game that introduces key building blocks of coding and provides an experimental setting in which we conduct evidence-based research to better understand the impacts of avatars and graphical content on students’ learning.
To disseminate results, we developed open-source materials supporting teaching and learning using our curriculum, which builds upon aspects of the nationally recognized Exploring Computer Science (ECS) curriculum and includes custom assessment instruments.
Through this project, we are Transforming the Educational Landscape with the belief that students can be agents of change in their own lives and the world, by bringing culture into the fabric of computing practice.
Alan Peterfreund
Thanks for the video. What evidence are you collecting on student impact?
Danielle Olson
Ph.D. Student
Hi Alan,
Activities were videotaped and audiotaped and all computer monitors were screen-captured. Avatar customization actions were logged by the AIRvatar system, a custom telemetry system to capture data about user avatar customization. All paper-based student materials were collected and photographed. In-game performance data and student survey data (regarding topics such as students’ dispositions toward computer science) were collected. Individual semi-structured clinical interviews were conducted roughly one month after the workshop.
Various qualitative research strategies from grounded theory methodology which were employed to analyze student discourse data. Let me know if this answers your question!
Brian Drayton
I am curious about the theoretical framing of your research — What makes you think that avatars can make a substantial contribution to students' interest and substantive engagement?
Danielle Olson
Ph.D. Student
Hi Brian,
Great question. I've copied a section from our latest publication (which will be published as a paper for the iRLN 2018 conference this June) below for your reference to answer your question. Additionally, I'd recommend looking into the extensive research on Virtual Identities / Blended Identities by D. Fox Harrell and Dominic Kao for additional studies that have demonstrated the effects of virtual identities on students' learning outcomes.
"The emphasis on avatar creation and customization within the workshops necessitates an understanding of the relationships between physical-world and virtual identities. Existing research on player-avatar relationships (PAR) is abundant, describing the human and non-human, physical and digital, material and immaterial parts which are “broken down and rebuilt” [10] throughout the process of avatar creation, customization, and use. A prior study on the relationships between underrepresented students in STEM and their avatars in learning games [11] characterized students’ perceptions of the construction and use of their avatars across three dimensions: (1) avatar appearance, with preferences ranging from everyday to extraordinary categories, (2) avatar ontological status, with perceptions ranging from first-person mirror representations to third-person external representations, and (3) avatar use, with uses ranging from deployment as instrumental tools to a means for imaginative identity play. Later studies have reinforced and broadened these three main dimensions more generally to PAR as a player’s (1) identification with their avatar, (2) attachment to their avatar, and (3) perception towards the avatar’s instrumentality [12][13], describing the degrees of self-similarity, affinity due to likeness, physical control, responsibility, and suspension of disbelief [14], and usefulness that players perceive their avatars to have [15].
Given the broad range of possibilities afforded by avatar creation and customization systems, extensive research has also been performed to better understand the motivations and behaviors of players in their selection of character traits for their avatar across artistic, psychological, and technological factors. In seeking to better understand the ways in which players’ physical and virtual world identities are blended through the personalization process, Harrell’s notion of blended identity [16] provides a mapping between aligned aspects of a player’s physical identity (e.g.: actions, characteristics, capabilities), the virtual identity (e.g.: technical system affordances and properties), and blended identity (i.e.: the playable character which is under the scope of user control)."
References from this excerpt:
10. Haraway, D.: Simians, cyborgs, and women: The reinvention of nature. Routledge, New
York (1991)
11. Harrell, D.F., Veeragoudar Harrell, S.: Imagination, Computation, and Self-Expression: Situated
Character and Avatar Mediated Identity. Leonardo Electronic Almanac. DAC 09: After
Media: Embodiment and Context. Vol. 17, no. 2, pp. 74-91 (2012)
12. Banks, J.: Object, Me, Symbiote, Other: A social typology of player-avatar relationships.
First Monday, vol. 20, no. 2 (2015)
13. Rehak, B.: Playing at being: Psychoanalysis and the avatar. pp. 103–127. In: Mark J.P. Wolf
and Bernard Perron (editors). The video game theory reader. Routledge, New York (2003)
14. Lewis, M.L., Weber, R., Bowman, N.D.: ‘They may be pixels, but they’re MY pixels’ Developing
a metric of character attachment in role-playing video games. CyberPsychology &
Behavior, volume 11, number 4, pp. 515–518 (2008). doi: 10.1089/cpb.2007.0137
15. Schultze, U., Leahy, M.M.: The avatar-self relationship: Enacting presence in Second Life.
ICIS 2009 Proceedings (2009). http://aisel.aisnet.org/icis2009/12.
16. Harrell, D.F., Lim, C.-U.: Reimagining the avatar dream. Communications of the ACM,
60(7), pp. 50–61 (2017). doi: 10.1145/3098342
Angie Kalthoff
Technology Integrationist
Are students able to create a video game around any issue they choose? Is there a framework to model types of issues such as the United Nations 17 Sustainable Development Goals?
What type of background information do students have coming into this workshop?
What happens with the game once a student has created and identifies that it is complete?
Danielle Olson
Ph.D. Student
Hi Angie,
Are students able to create a video game around any issue they choose? Is there a framework to model types of issues such as the United Nations 17 Sustainable Development Goals?
Yes, students were led in a design thinking exercise in which they were asked to independently / silently write ideas on sticky notes about topics / themes / issues of their choosing. Once all notes were collected, the themes and ideas were clustered into similar topics and students chose to form teams around these ideas. The sustainable development goals were not explicitly noted in the exercise, but many related topics (from education equity to improving police-public relations) of consequence were ideated by students.
What type of background information do students have coming into this workshop?
Students were assigned 4 readings / videos with accompanying written responses as homework related to the 4 topics addressed in discussion during the workshops: Stereotypes in CS/Engineering, Social Media and Social Connections, Online Privacy and Surveillance, and Monetization and Advertising. These reading ranged from academic papers, to pop science and tech policy articles, and the videos were YouTube clips of various representations of computer scientists (from everyday to extraordinary).
What happens with the game once a student has created and identifies that it is complete?
The students participated in an iterative design process. First they did paper prototypes of the games, then digitally implemented them, then had their peers give feedback on their games, then they finalized these creations by end of workshop. Once these were complete, we held a student presentation session in which the student teams presented to their peers and also reflected on their learnings and thematic ideas from their creations.
Angie Kalthoff
Technology Integrationist
Thank you for your response!
Are you able to share the 4 readings/video that students were required to complete? I would love to explore them as well.
Danielle Olson
Ph.D. Student
For sure! Please find links for the 4 topics below:
Advertising and Monetization:
Stereotypes about computer scientists:
Social media and social connections:
Privacy, surveillance, and self-expression:
Karthik Ramani
Donald W. Feddersen Professor of Mechanical Engineering
I can see why use of avatars can create a personal engagement for the learner. But in terms of best practices was not able to distill the principles for doing so. Could you shed some light into this part as to why avatar use can trigger excitement? Is this true of all personality types? Also some comments of going beyond was made such as HCI - but no other examples were provided. Any more? And Why only restrict to make what exists? (comment at towards the later part).
Danielle Olson
Ph.D. Student
Hi Karthik,
The following fundamental concepts from the Exploring Computer Science (ECS) curriculum were incorporated into the workshop:
Beyond these CS concepts, students engaged in homework/readings/writings and discussions to critically reflect on these topics of technology and society:
Regarding your question on principles for using avatars in immersive learning: please find the below academic papers published by our research team which validate how avatars specifically trigger learning outcomes and engagements for students:
A few places where principles can be found are: Abstract vs. Likeness Avatars: Toward Evaluating the Impacts of Virtual Identities on STEM Learning [ paper], Dominic Kao, D. Fox Harrell, Foundations of Digital Games (FDG 2015), Pacific Grove, CA, June 22 – 25, 2015. Role Model Avatars: Exploring the Use of Role Model Avatars in Educational Games [ paper], Dominic Kao, D. Fox Harrell, Experimental AI in Games (EXAG 2015), Santa Cruz, CA, November 14 – 15, 2015 Exploring the Impact of Role Model Avatars on Game Experience in Educational Games [ paper], Dominic Kao, D. Fox Harrell, Computer-Human Interaction in Play (CHI PLAY 2015), London, England, Oct 5 – 7, 2015. Toward Avatar Models to Enhance Performance and Engagement in Educational Games [ paper], Dominic Kao, D. Fox Harrell, Computational Intelligence in Games (CIG 2015), Tainan, Taiwan, August 31 – 2, 2015. A good summary of principles is in: Researching and Developing the Impacts of Virtual Identity on Computational Learning Environments, Dominic Kao, Ph.D. Dissertation, 2018. An excerpt (http://people.csail.mit.edu/dkao/pdf/dominic-thesis-excerpt.pdf):"1. Using avatars that resemble users when they are doing well, and appear more
minimally or abstractly otherwise, is encouraged whenever possible. The research
in this dissertation, which defines these as successful likeness avatars, has shown
that they result in improved user performance and engagement [13]. For example,
applied to a mail client or a social network like Facebook, your icon would change
between a likeness of yourself or abstract depending on the positivity of your news
feed, or a message you received. The essence of this principle is selectively promoting
detachment and identification at key moments of the digital experience.
2. Using avatars that resemble role models is encouraged whenever possible. The
research in this dissertation has shown that role model avatars increase both the engagement
and performance of users [10–12, 17]. For example, playing as an admired
and positively influential scientist, politician, business person, artist, or doctor depending
on context. The criteria for an effective role model is perceived competence,
similarity, and success, therefore role models should represent successful figures with
demographic overlaps with users.
3. Use embellishment with trade-offs in mind. The research in this dissertation has
shown that embellishment increases engagement, but decreases performance and
self-efficacy [16]. For example, in an educational context, embellishment can be
reduced to promote performance and self-efficacy, while in an entertainment context
embellishment can be used more liberally.
4. Using positive or neutral encouragement is encouraged whenever possible. The
research in this dissertation has shown that positive (e.g., “Keep it up!”, “Don’t give
up!”, “You’re almost there”) and neutral (e.g., “You are doing standard work”, “You’re
doing average”, “You’re doing typically”) encouragement text increases engagement
[15]. For example, encouragement text can be spoken by a game character, or simply
appear at the bottom of the screen periodically.
5. Promoting avatar identification is encouraged whenever possible. The research in
this dissertation has shown that avatar identification promotes higher engagement,
self-efficacy, time spent, and even quality of created artifacts [17]. For example,
giving users the ability to customize their avatars is one simple way of increasing
identification."
Dissertation References:
[10] D. Kao and D. F. Harrell. Exploring the Impact of Role Model Avatars on Game
Experience in Educational Games. The ACM SIGCHI Annual Symposium on Computer-
Human Interaction in Play (CHI PLAY), 2015.
[11] D. Kao and D. F. Harrell. Exploring the Use of Role Model Avatars in Educational
Games. In Proceedings of the AIIDE Workshop on Experimental AI in Games, colocated
with Artificial Intelligence in Interactive Digital Entertainment, 2015.
[12] D. Kao and D. F. Harrell. Toward Understanding the Impacts of Role Model Avatars
on Engagement in Computer Science Learning. In The annual meeting of the American
Educational Research Association (AERA), 2016.
[13] D. Kao and D. F. Harrell. Exploring the Effects of Dynamic Avatars on Performance
and Engagement in Educational Games. In Games+Learning+Society (GLS 2016),
2016.
[14] D. Kao and D. F. Harrell. Exploring the Impact of Avatar Color on Game Experience
in Educational Games. Proceedings of the 34th Annual ACM Conference Extended
Abstracts on Human Factors in Computing Systems (CHI 2016), 2016.
[15] D. Kao and D. F. Harrell. Exploring the Effects of Encouragement in Educational
Games. Proceedings of the 34th Annual ACM Conference Extended Abstracts on
Human Factors in Computing Systems (CHI 2016), 2016.
[16] D. Kao and D. F. Harrell. Toward Understanding the Impact of Visual Themes and
Embellishment on Performance, Engagement, and Self-Efficacy in Educational Games.
The annual meeting of the American Educational Research Association (AERA), 2017.
[17] D. Kao and D. F. Harrell. The Effects of Badges and Avatar Identification on Play and Making in Educational Games. In CHI, 2018.
Thanks,
Danielle
Further posting is closed as the showcase has ended.