Welcome to Dataville

Welcome to Dataville is a game designed at NYU, in collabortion with faculty from NYU Steinhardt, to help teach hypothesis testing. You play the role of a city official running for mayor. You are tasked with making various improvements aournd town using sound statistical decision making. Successful problem solving improves your ratings in credibility, popularity, and competence. With high enough approval ratings, you can complete the game and earn the status of mayor of Dataville.
I took on the role of lead developer. I was also involved in the design stages, especially working on the contextual help and scaffolding mechanisms. Addtionally, I created some graphics, integrated it with AWS Lambda, DynamoDB, and SES (Simple Email Service). I performed beta testing with student focus groups. Students were provided pizza. Welcome to Dataville was built in Unity for Windows and Mac.

Educational Scaffolding

When designing Welcome to Dataville we made educational scaffolding a top priority. Students start the game with only basic knowledge of statistical terms. On easy tasks they are presented with step-by-step instruction, labeling of terms, and actual answers to problems which can be revealed as needed. As they progress through the game the scaffolding is removed. Medium questions no longer provide answers, but they still support decision making and provide definitions. Once a student reaches the hard questions they should be able to independently solve the problem. As the difficulty progresses the benefit to their rankings also increases, so students are naturally encouraged to solve difficult problems at their own pace.

Customization / Feedback

Scenarios also provide randomized data so that students will never encounter the exact same question. The data presented in each problem is saved, so it can be referred back to if a student has questions or concerns. Players are also provided with feedback based on their performance, recommending different categories or suggesting they go back and work with an easier mode if they are struggling. Anonymized data is stored so professors can get an overall idea of where a class is struggling, hopefully informing their lessons.