Information Technology Reference
In-Depth Information
the player must obtain items from all three of the other players. Put differently,
we expected that players would be more likely to experience greater conflict when
pursuing
Quest 4
than when pursuing
Quest 2
.
To test our assumptions regarding quest design and in order to model the conflict
escalation and deescalation process, we collect the following three metrics from
players: experienced conflict intensity, emotional state, and current feelings towards
other players following major game events (e.g. completion of quests, trading
success and failure, stealing, message posting, and spreading rumours).
Although the game can continue indefinitely, the game has a number of badges
that players can collect. These badges both serve the purpose of introducing to
players to the possibilities of the game world, while also articulating play goals
for them to work towards, some of which concern game progress, others of which
concern conflict learning progress. Examples of badges include “Rude surprise”,
which is earned by a player when another player leaves a bad comment about her,
“Change of heart”, which is earned when a player registers a significant change
of feelings about another player, “There's No “I” in Team”, which is earned after
successfully completing a trade with three players during a single quest, and “Wise
Words to Live By”, for giving helpful advice to other players during the reflective
debrief conversation.
21.4
Experimental Protocol and Data Collection
In order to test the game's ability to induce the escalation and deescalation of
conflict, as well as reflective learning of conflict resolution strategies, we conducted
a game user study. The user study also served the purpose of collecting data for
training the computational user models enabling the game's scenario adaptation.
The study was run at a Portuguese school with 32 (20 male) children between the
ages of 10 and 12 who played 21 game sessions in total. The maximum duration of
each play session was 15 min. All sessions consisted of groups of four players who
were classmates. All groups consisted of children of the same gender. Choosing
classmates with prior relations meant that the actions and evaluations in the game
plausibly would be influenced by these relations. However, since this would be the
natural use case for the game in a school setting, we consider this advantageous,
rather than detrimental, to the user study.
Every participant played multiple game sessions (2-4). In total, we collected
2,126 reports about conflict intensity, emotions, and judgement about the other
players across the 21 game sessions. In order to establish a baseline for assessing
the short-term impact of interacting with the game, the students were asked to
report their emotions and judgements about the other players at the very start of
the session. These ratings were updated whenever key events occurred in the game
environment. Events included rumoring, message posting, trading, and stealing
actions. The conflict intensity in the beginning of the game was assumed to be 0
Search WWH ::
Custom Search