Speakers: Jan Plass, Ken Perlin and Katherine Isbister from the Games for Learning Institute

When it comes to using things like games for education, producers try things without knowing if they worked or not.  Goal is not to make one killer game, but to make games that are scientific instruments to study what works and what doesn't work.  They make lots of mini games.

Factory Actor: several variants on one game, variables to change to see how players learn prime factorization.  Better to compete, or to have 2 players work together?  Also Super Transformation, Prime Beef, EcoSim  See http://g4li.org

Unmouse Pad, put hand right on it, pressure sensitive.  Allows for affective computing.  Less expensive than a tablet. Posterior (butt) sensor in chair to measure posture -> interest, slouching = boredom, leaning in = interest

Parts of a learning game: 1) head, player's understanding  2) game mechanic, rules of play 3) aesthetic design, graphics & sound  4) narrative drive, moves story forward  5) extrinsic rewards, points, ranking  6) intrinsic rewards, improving skills **most important part** this moves you up, whereas #4 moves you forward.  #6 is closely tied to Csikszentmihalyi's concept of Flow (see my notes from Boston GameLoop 2009)

Design Process: Interview and observe kids and game designers who make learning games (presented at CHI 2010 Designing Games for Learning)  Findings: Deep content not 'bolted on', with innovative mechanics, exploration of systems, engage emotions.  Got list of things teachers find difficult to convey with traditional classroom teaching methods.

Research: Study impact of player characteristics on outcomes.  Design patterns that would be reusable.  Run experiments to compare different methods.  Playtech by Colleen Macklin.  Rapunsel: teaches girls to program (My note: Sounds similar to Alice at CMU) outcome was designing an avatar.  Concept of self as programmer, and self-esteem increased, thinks of self more as scientist  Simulations Research: teaching HS chemistry, iconic rep of key info, sim exploration v. direct instruction, sim efficacy, engagement & graphing skills Self exploration is better than just feeding facts (contrary to other published research).  Emotion Research: Positive emotions increase comprehension and transfer.  Eye tracking research: expert/novice differences.

Design patterns for edu games: genre-specific, topic-specific, constructing things is fun, strong narratives provide sufficient incentive (like Prof Layton), games can be engaging without visuals (Tribalwars), games can be engaging and addicitive without necessarily being fun.  The more intrinsic motivation, the less extrinsic is needed.

Engagement is much better in their lab than at the after school centers.  Triangulation of self report, behavior and bio-metrics.  Embedded assessment in the game.  What does it mean, within the game, to have something learned.

More prepared student vs less prepared student trajectories.  Must adapt to both.