Team

• Principal investigator / project leader: Dmitriy Babichenko, PhD
• Game designer / developer: Patrick Healy
• Research assistant: Kylie Dougherty
• Domain expert: Cynthia Danford, PhD

An Action/Adventure Game to Teach Children About Nutrition

The worldwide prevalence of childhood obesity has increased greatly during the past several decades.  Numerous studies show that the increasing occurrence of disorders such as type 2 diabetes in children is likely a consequence of this obesity epidemic [1]. In recent years a number of methodological approaches and interventions have attempted to stem the obesity epidemic. These interventions usually utilize one or more of the following approaches: “dietary, physical activity,  healthy  lifestyle  education,  and/or  parental  involvement [2], [3]”. 

Most healthy lifestyle education interventions are school-based and are either presented as part of health-related courses or as school-based afterschool programs.  In general, these interventions show that children from the intervention group were “significantly more likely than comparison students to choose more healthful snacks when given the opportunity [4], [5].”

One of the major drawbacks of the aforementioned interventions is sustainability.  In many cases the intervention ends once the academic study is over. Even when interventions become part of the school curricula they require significant resources to train faculty, create educational materials, and time investment to continuously reinforce the importance of correct nutritional choices. 

Another important shortcoming is the content delivery itself.  Most of the reviewed studies [3] delivered healthy lifestyle education intervention via traditional content delivery methods, such as lectures and videos.  

Gamification is the use of game-like strategies and elements in a non-game setting [6], [7]. Its incorporation into educational disciplines has shown the ability to increase student engagement and motivation, potentially resulting in better student learning outcomes [6]–[10]. 

Among the most important factors pertaining to the successful use of game element integration is learner satisfaction with the gamified learning environment [11]. Interventions require gamification to align with learning goals already in place, as well as to provide near-immediate feedback to students [10].  In doing so, game elements, such as clearly defined objectives and rules, must be addressed in regards to the dynamics of a student’s interaction, but can also be customized to the learning objectives and the desired learning outcomes [11]. A clear framework for the rewards and penalties that a student can obtain for classroom attendance and overall performance, including completion of tasks that may have garnered resistant behavior in the past, can actually improve measures of learning across the board [12]. When correctly implemented, gamified learning environments show higher rates of motivation and engagement compared to a traditional didactic classroom, thus leading toward positive learning outcomes and improvement in knowledge acquisition and retention [13]–[15]. 

Multiple recent studies that investigated the use of games as learning tools have indicated that game play can be responsible for significant increases in knowledge and cognitive performance [16]–[18]. Studies also report that trial and error strategies included in-game mechanics are more conducive to knowledge acquisition and retention than traditional classroom environments [16], [19].

The acquisition of knowledge from playing video games is not surprising given that video games are intrinsically laden with the same attributes that are congruent with learning such as improved coordination (von Salisch, Oppl, & Astrid, 2006) interactivity (Klimmt & Hartmann, 2006), challenge (Vorderer, Hartmann, & Klimmt, 2003), and reward. Interactivity or the simultaneous open feedback loop of communication between player and video game allows a continuous process of trial and error and problem solving, which is an important component of knowledge acquisition. 

There are quite a few games dedicated to nutrition education that have been used in numerous studies and have been validated in the context of the acquisition of nutrition knowledge.  Games such as Lunch Crunch, Food Fury, and Juice Jumble from Playnormous [19], Nutritional Bingo [20], [21], and Kaledo [22] have all shown to improve knowledge and understanding of nutritional choices in both children and adult populations.  Naturally, the question that comes to mind is, “Why waste resources to develop yet another nutrition education game?”

Current nutrition education games described in the literature have a number of shortcomings that may hinder scalability and widespread adoption. Games such as Nutritional Bingo [20], [21], and Kaledo [22] are tabletop games and thus are limited to a small number of players.  Moreover, these games require participation and guidance from a gamemaster – an educator or a researcher who “acts as an organizer, officiant for regarding rules, arbitrator and moderator” during game sessions [23].  Digital games designed to promote nutrition knowledge and a healthy lifestyle [24] are generally designed as edutainment experiences [25]–[27].  In other words, the design of educational content takes priority over the entertainment part.  Unfortunately, to the best of our knowledge, there is no concrete empirical evidence in the scientific literature that indicates which game design approach results in better learning outcomes – edutainment experiences or entertainment games with educational components.  

To address some of the shortcomings of existing interventions, we designed MountainQuest, a multi-level action/adventure digital game that teaches basic nutrition knowledge to elementary and intermediate school-age children through a series of decisions built into the game narrative and mechanics.  

Our long-term academic intent is to explore the following research questions: 

1. RQ1: To what degree is it possible to improve the acquisition of knowledge of nutritional choices in children ages 8-12 using a digital interactive game compared to an educational game that has no entertainment components?
2. RQ2: To what degree does information delivery through a digital interactive game improve the retention of knowledge of nutritional choices in children ages 8-12 compared to an educational game that has no entertainment components?

Design Choices

MountainQuest 3.0’s individual levels came in two types: linear railroads and collect-a-thons. In linear railroads, players are given clear goals, indicated by a guiding arrow on the ground in front of the player character, which they simply must navigate the environment to reach. In collect-a-thons, players must freely navigate the environment to find a number of hidden objects in the form of sheep that the trolls have stolen from the village. Later collect-a-thon levels introduce a rope swinging mechanic that allows the player to swing on a rope that they may attach to any surface in their environment, similar to the web-swinging mechanics of Spider-Man 2 (2004).

The game is significantly focused on managing an inventory of food. Sacks of food are scattered around the game environment and if the player reaches them, food will be added to their inventory. This food can either be eaten or thrown. Eating food will affect the player’s metrics of hunger, thirst, and energy. These three metrics are always shown on the heads-up display and affect player movement. If the hunger or thirst metrics reach a maximum value or the energy reaches its minimum, the player’s movement speed is slowed to a crawl. Otherwise, the player’s movement speed is scaled to the energy metric. These metrics will change over time either through exercise (player motion), which will decrease energy and increase both hunger and thirst or through eating food, which will affect the three stats in various ways depending on features of the specific food the player has chosen to consume (Figure 1). 

Figure 1: Hunger, thirst, and energy meters are shown at the upper left corner, along with the player’s current objective. The food selection menu is at the bottom right corner.

Throwing food is used for the game’s “combat” mechanic. At times, trolls will attempt to steal all of the player’s food. If a troll reaches the player, they will reach a fail state and have to restart at a checkpoint. To avoid this, the player can throw single food items at the trolls, which will disable them if the item is processed food or otherwise deemed unhealthy. This forces the player to make on-the-fly decisions about which food items they should consume and which they should throw away (Figure 2). 

Figure 2: The player has just thrown a taco at one of the two trolls in front of them. The troll fell asleep, indicated by the letter ‘Z’ above their head.

As part of MountainQuest 3.0 redesign, we also updated the feedback mechanic.  Instead of popup messages, the player receives nutrition tips through the game’s narrative, conveyed primarily in conversation between the player character and non-player characters through a text and voice dialog system. The bulk of nutrition information is conveyed by a bird companion character that interacts with the player regularly, providing in-game guidance, nutrition tips, as well as feedback on poorly made nutrition-related choices and praise for healthy food choices (Figure 3). Table 1 shows several examples of nutrition tips provided by the bird.

Figure 3: The bird character in a conversation about processed foods with the player.

Table 1: Examples of nutrition tips provided by the in-game bird avatar

Nutrition Tips

You should try to make up as much of your diets as possible from totally unprocessed foods! Unaltered fruit, vegetables, eggs, and meat are almost always healthier than a packaged version.

It’s not always easy to eat only unprocessed foods, but at least try to limit how much processed foods you eat. Processed boxed pasta isn’t the healthiest food, but if you eat it with some whole fruits and vegetables, it can still be a healthy meal!

Try to totally avoid heavily processed foods – things like cookies, chips, and cereals. These things can taste pretty great but you don’t get much nutrients from them, except a whole lot of sugar.

Discussion and Future Work

It is worth mentioning that the design and development of MountainQuest involved participation from undergraduate and graduate students across multiple disciplines, including computer science, information science, music, English, pharmacy, nursing, and sociology. Moreover, as part of the participatory design effort, we worked with the University of Pittsburgh community engagement centers and several Pittsburgh area schools to run game design workshops with intermediate and middle school students.  This approach provided an interesting opportunity to teach intermediate school students about computational thinking and computer programming constructs through game design activities, such as coming up with algorithms to get a game character from point A to point B, or estimating hunger and thirst levels based on player activities.

One of the most important design lessons that we learned as a team is that children do not understand (or do not accept) the concept of a prototype.  While in our past projects we constantly relied on wireframing and low fidelity prototypes to elicit feedback from end users, working with children taught us to bring fairly polished game scenes and levels to playtests.  Even the smallest things would break the game play and shit the children’s attention from the task at hand.  For example, while playtesting MountainQuest 2.0, we did not have a chance to fully implement the algorithm which would allow troll non-player characters (NPCs) to follow and attack the player.  In the version that we brought to the playtest, the troll moved sideways in a somewhat jerky motion and walked through “physical” obstacles. Every playtest participant (five subjects) stopped the gameplay to comment on the NPCs movements.  When we told the participants that the version of the game that they were playing was a prototype and that they should continue to interact with the NPC according to game rules, ALL subjects spent between 3 and 7 minutes trying to get the NPC to walk through the mountain instead of following the playtest protocol.  We observed similar behavior with every playtest group and with every bug that they found in every version of the game.

The second lesson boils down to the fact that children are spoiled by the wide array of highly polished games that are available on Android, Apple stores, as well as on the Steam platform. Designing meaningful serious and / or transformational games that could potentially compete for children’s time is extremely difficult.  While we do not have a good insight into how to address this issue without a game studio-level budget, one possible approach is to work with schools and community centers to introduce these types of educational games as part of the school curriculum.

Acknowledgments

We would like to thank the University of Pittsburgh Pitt Seed Projects program for providing partial funding for this program.  We would also like to thank Dylan Fuhr, Max Edelstein, Jacob Buchinsky and Austin Frownfelter for developing the initial prototypes of the game, Karen Brown for providing design insights and writing music score for MountainQuest 2.0, Dr. Victoria Grieve for her constant help with game design tasks, and Dr. Ravi Patel for keeping us on track with the educational relevance of the game.  Last, but not least, we would like to thank Drs. Kayla Booth, Victoria Grieve, and Ravi Patel for being co-investigators on this project – without your help and support we would have never been able to obtain the funding and get the project to its current state!

References