Designing Learning Games to Maximize Engagement

In recent research appearing in Simulation & Gaming, Rodrigo1 explores the cognitive-affective states students experience over the course of a learning game by observing 30 Filipino students at 12 intervals playing a pre-algebra game called Math Blaster (for Ages 9-12).  Most surprisingly from this research, it was discovered that the state of confusion leads to student engagement.

Math Blaster Ages 9-12 by Knowledge Adventure

Evidence for seven general states was found:

1. Boredom
2. Confusion
3. Delight
4. Engagement
5. Frustration
6. Surprise
7. Neutral (none of the above)

For learning to occur, we generally want students to maintain the engagement state, where they are immersed and focused on the material.  In Math Blaster, engagement was most common, followed by boredom, followed by delight.

But the states themselves are not as interesting as the transitions between them.  Which states precede learning?  Which states precede other states?

Most critically, boredom begets boredom.  Once a student entered the boredom state, it was somewhat probable that they would stay bored, and highly probably that they would not transition to engagement.

But most interestingly, confusion begets engagement.  If a student becomes confused, it is highly probably that they will transition to an engagement state.  The author claims that this follows from previous research suggesting a correlation between confusion and achievement, but it is not explored deeply in the paper.

I personally believe this is more closely related to the experience of accomplishment – the student is confused, overcomes that confusion, and then is motivated to move forward by that victory.  This seems related to the psychological theory of motivation by goal setting.  Accomplishing goals motivates us to accomplish other goals.

Generally, this speaks to an important game design principle for learning games – they can’t be too easy.  If we consider the engagement state to be where most learning occurs, we must design learning games to push users to that state.  There are generally two options to go about this: 1) start easy and gradually ramp up the difficulty or 2) start difficult (but not too difficult!) and provide the player with advice and guidance to improve their skills to that level.  From this research, it appears that the second approach is preferable – at least with children.  Further research should examine this with older populations to see if these principles generalize to adults and college students.

1. Rodrigo, M. (2010). Dynamics of student cognitive-affective transitions during a mathematics game. Simulation & Gaming, 42 (1), 85-99. doi: 10.1177/1046878110361513 [↩]