Pr

1
Impacts of Awareness Tools on Mutual Modelling
in a Collaborative Video-Game
Nicolas Nova CRAFT/EPFL CRIWG_2003
2
Overview

• The collaboration process and the concept of
  awareness
• Mediated Collaboration and its drawbacks
• Aim of the research and hypotheses
• Methodology
• Results
• Additional analyses
• Conclusion/limits/future work

3
Collaboration

• Grounding awareness acknowledgement
• Attribution/Mutual modelling
• Group transactive memory

4
Awareness

• Awareness is an understanding of the activities
  of others, which provides a context for your
  activity  Dourish Belloti
• Who ? What ? Where ? How ? When ?
• Synchronous and asynchronous awareness
• Awareness for what ?

5
Mediated collaboration
Simple collaboration
Mediated Collaboration Lack of
information Concerning awareness
6
Awareness tools

• In order to support awareness in mediated
  collaboration ? Awareness Tools (AT)
• Examples chat, radar in counterstrike, social
  navigation, etc.
• Lack of interest for the study of social and
  cognitive impacts or the use of those AT. Juste
  few studies about their usability.

7
Research question

• The cognitive effects of using Awareness Tools on
  Mutual Modelling in collaborative problem solving
• 5 hypotheses

Awareness Tool
Mutual Modelling
Performance
8
Methodology SpaceMiners
9
Methodology awareness tool
Awareness Tool localization partners gaze
10
Methodology procedure and participants

• Participants gamers ?
• 36 persons 9 pairs in each condition (with AT
  and without)
• Partners within a pair do not know each other
• Procedure
• Questionnaires
• 2 houres max.
• Its hard to play !

11
Methodology questionnaires

• Two multiple-choice questionnaires what do you
  intend to do at the moment ? What do you think
  you partner intends to do ? Guiding ones
  partner, trying to shoot, trying to establish a
  common strategy, adjusting a shoot
• ? Mutual modelling indexes
• Micro-evaluation, related to a particular event.
• Paper-based questionnaire players were asked
  whether they could guess what their partner was
  doing during the whole game.

12
Results about Spaceminers

• Almost all had fun (28/36).
• 3 pairs less than 1h30.
• No major problems.
• We cancelled 2 experiments

13
Results hypothesis 1

• Significative effect of the AT on performance
  (score) F4.84, p 0.043.

14
Results hypothesis 2
No effect of the AT on Mutual Modelling accuracy
(p .889).
15
Results hypothesis 3

• Trend of effect of time on Mutual Modelling
  accuracy, but statistically not significant (p
  .084).

16
Results post-hoc analyses

• Analyses on contrasted groups (7 best MM or
  score), t student.
• No effect of score on MM (p 0.078).
• Effect of MM on score (p 0.045).
• AT does not help a lot good pairs to improve
  the accuracy of their MM

17
Additional results

• Pairs with AT better self-estimation of their
  partners comprehension
• No correlation between A?B and B?A.
• No statistical differences in the number of
  actions performed.
• HOWEVER pairs with AT who spent more time in
  camera view had a more accurate MM.

18
So what ?

• Effect of AT on performance and on MM only if
  they use the AT.
• So Awareness Tools or not ?

19
Limits et critics

• Low number of participants, differences between
  pairs (time spent)
• Measure of mutual modelling
• Our Awareness Tool
• ! Grounding awareness acknowledgement.

20
Whats next ?

• My phD ? focus on socio-cognitive impacts of
  location-based awareness (mobile devices)
• Links between space, cognition and collaboration
• CSCL