Discussion

            Through the results it has been clearly proven that there is a difference in quality of communication experienced when a digital avatar is implemented in chat communication.

However the ANOVA repeated measures significance tests failed to prove that marginal difference between direct and indirect avatar gaze conditions were significant at the 0.05 level. This can be attributed to the following reasons

1.                          Small sample population of 34. Because of this, the sample most likely was not a normal bell curve. However it is interesting to note that when comparing direct and indirect gaze, the p value for partner evaluation was only greater than 0.05 by a value of 0.044. This suggests that redoing the experiment with a much larger sample size may have promising results.

2.                          Participant external conditions. Although every procedural precaution has been taken to make the virtual experience as uniform as possible, there are many variables of reality that may have confounded the results. Since the participant in reality was not undergoing the sessions in a real closed environment, and due to differences in computer hardware and internet speed, the virtual experience of direct and indirect avatar gaze may not have been experienced properly.

3.                          Poor representation of direct gaze behaviour. Although the direct gaze was meant to represent real human behaviour, it may have not been designed to meet such expectations.

4.                          Chat history location. In mouse-look mode the chat history is shown on the bottom left corner of the screen. If the chat history is moved above the head, it maybe possible to turn the participant’s focus towards the head movements of a direct gaze avatar.