Gaming

Research Suggests Students Learn More When Working Together in Virtual Reality Games

• By Michael Hart
• 06/22/16

With the help of a virtual reality game their professor created, some students at Cornell University have gathered evidence that interacting with other people in a game-playing atmosphere may help them learn more.

Using a language-learning game, "Crystalize," created by Cornell Assistant Professor of Computer Science Erik Andersen, a group of students conducted a study with two different groups in which players, using an avatar, take a virtual tour of Japan.

In the game, which is intended to help students learn Japanese — although it is designed to teach other languages as well — players go on "quests" to learn new words by watching game characters talk to each other. For instance, as one character walks away, another that is left behind says, "Sayonara." The hope is that the player understands "Sayonara" means "goodbye." At that point, the player can drag the word from a speech balloon into an inventory of terms that can later be used to construct sentences.

Naturally, players collect points by completing quests, as is so often the case in role-playing games.

In the research project led by Ph.D. candidate Gabriel Culbertson, 48 students were recruited to play two versions of the game. In one group, students were connected via a chat interface with another player who could, if they wanted, offer advice on how to play. The second group played a version of the game in which they were definitely required to collaborate on quests.

The research group found the students in the second so-called “high-interdependence” group spent more time communicating and, as a consequence, learned more words.

The research then expanded to a larger group of 186 Reddit users who were learning Japanese. After reviewing gameplay logs, interviews and Reddit posts, they found that those who spent the most time engaged in the game learned more new words and phrases.

The Cornell research team presented its research results at the Association for Computing Machinery Conference on Human-Computer Interaction in May in San Jose, CA.

The game's designers plan to look for ways to create longer-term engagement with the game since learning a language is a long-term process.