The Learning Companion asks which ring you would like to move in trying to solve a puzzle. The current version may look clunky, but as the Learning Companion becomes more advanced, its appearance will become more sophisticated—and more human.           

Affective Learning Companions:  Promoting Metacognitive Strategies for Perseverance through Failure


Winslow Burleson, Rosalind Picard

Affective Computing Group

MIT Media Laboratory


Developing learning experiences that facilitate self-actualization and creativity is among the most important goals of our society in preparation for the future. To facilitate deep understanding of a new concept -- to facilitate learning -- learners must have the opportunity to develop multiple and flexible perspectives. The process of becoming an expert involves failure, understanding failure, and the motivation to move onward.  Meta-cognitive awareness and personal strategies can play a role in developing an individual’s ability to persevere through failure and combat other diluting influences. This research centers upon the development of a theory for using affective sensing and appropriate relational agent interactions to support learning and meta-cognitive strategies for perseverance through failure.  We are investigating, designing, building, and evaluating Relational Agents that may act as intelligent tutors, virtual peers, or a group of virtual friends to support learning, creativity, playful imagination, motivation, and pursue the development of meta-cognitive skills that persist beyond interaction with the technology.


Failure is important to learning and instrumental to the development of multiple points of view required for deep understanding.  Failure and repeated failure can also have a negative impact on motivation, affect, and learning.   Therefore, learners’ response to failure is very important to their continued learning.   Alice Isen has found overwhelming evidence that mild positive affect improves negotiation processes and outcomes; promotes generosity and social responsibility; self-efficacy; motivation toward accomplishment; openness and flexible manipulation of new information. “Positive affect is a source of human strength… promoting thinking that is not only efficient, but also careful, open-minded and thorough.” (Isen, 2002) At the same time it is important to realize that the staying power of negative affect tends to outweigh the more transient experience of positive affect.  This is a phenomenon known as “negative asymmetry.”(Giuseppe, Brass, 2003)  Unfortunately for the purposes of motivating learners this negative asymmetry means that negative affect experienced from failure will persist disproportionately to the positive affect experienced from success.  Educators and innovators must try extra hard to create motivating learning environments which celebrate achievement and provide sustaining inquiry opportunity at times of failure.  The role of failure in learning and the use of thinking and technological strategies to support perseverance through failure is our focus.


Download a Video: ALC MOVIE.avi (51MB)



W. Burleson and R. W. Picard (2004), "Affective Agents: Sustaining Motivation to Learn Through Failure and a State of Stuck" Social and Emotional Intelligence in Learning Environments Workshop In conjunction with the 7th International Conference on Intelligent Tutoring Systems, Maceio - Alagoas, Brasil, August 31st, 2004. PDF


W. Burleson, R. W. Picard, K. Perlin and J. Lippincott (2004) "A Platform for Affective Agent Research," Workshop on Empathetic Agents, International Conference on Autonomous Agents and Multiagent Systems, Columbia University, New York, NY, July 2004. PDF


Press Coverage:


For more information:

Winslow Burleson’s Home Page

Affective Computing Group

Prior Work on Affective Learning Companions by the Affective Computing Group


NSF LogoFunded by the National Science Foundation

This material is based upon work supported by the National Science Foundation under Grant No. 0087768.  Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.