Research

Research

Research Results

Research Results

Research using the Virtual Violin Game

During 2015 a trial of the Virtual Violin Game was undertaken as part of Marcel Trussell-Cullen's PHD at th University of Melbourne. 12 students between the ages of 6 and 13 took part in the project. The trial consisted of two stages. The initial stage refered to as the recording stage the students recorded their lessons and practices and recieved no interactive feedback. The second stage referred to as the interactive stage involved use of the Virtual Violin Game where students receive real time interactive feedback about 9 positional parameters of violin playing as well as sound feedback consisting of volume and pitch accuracy data. Students recieved points based upon an algorithm calculated from the interactive data and cumulative total is refered to as a total score.

The Total Score of the students averaged less than zero during both the lessons (-0.79) and the practices (-19.81) during the recording stage. The recorded lesson results were higher than the practice results. During the interactive stage the average total score was positive during both the lessons (4.94) and the practices (15.85). The practice results are greater than the lesson results but still below the average result of the expert performers (41.9). Seven of the twelve participants had a positive average total score during the interactive stage of the project (37.4, 25.1, 18.8, 17.5, 14.0, 8.6, and 3.0) with two of the student participants achieving an average score that was comparable to the expert players. During the recording stage just three of the participants (13.1, 7.7, and 1.4) achieved an average total score that was positive. The top average score during this phase was significantly lower than the expert players’ average scores (41.9).

During the recording stage five students had positive improvement in their average total scores (27.5, 16.4, 7.8, 4.5, 4.0) but during the interactive stage of the trial eight of the twelve students demonstrated positive improvement of their average total score (94.8, 86.8, 38.6, 26.3, 13.3, 11.4, 10.1, 8.0). Furthermore the magnitude of the improvement of these students is significantly larger than in the recording stage. During the lessons of the recording stage four of the students showed positive improvement of their total score (33.2, 14.9, 4.5, and 4.0). In contrast during the lessons of the interactive stage seven students demonstrated positive improvement of their average total score (101.5, 40.2, 39.5, 16.5, 13.3, 11.4, and 8.0) and once again the magnitude of the improvement is significantly larger in the interactive stage. During the practices the results are opposite to the lessons. In the practices although there are the same number of students (6) whose average total scores improve or stay the same during the recording stage the average improvement of total scores (91.0, 42.7, 32.5, 28.6, 10.1, 8.4) is greater in magnitude than the interactive stages average improvement of the total scores (25.2, 14.7, 5.6, 2.7, 0.0, 0.0).

Pearson Correlations were taken with the improvement values of the average total scores during the recording and interactive stages. During the recording stage there were no Pearson correlations above 0.4 or below – 0.4 with the improvement of the player total score averages. During the interactive stage the results showed the strongest correlation between the total number of sessions with the game outside of the lesson (0.71). There were further significant correlations between the average improvement of the total score and the number of game sessions (0.70), the highest score (0.62), the use of the game (0.53) and whether the student had private lessons (0.46). There was also a strong negative correlation with the average improvement of the total score and the male students. (-0.45).

A paired-samples t-test was conducted to compare the total score average in the recording stage and the interactive stage. There was a significant difference in the average total scores for the recording stage (M= -10.3, SD=16.45) and the average total scores for the interactive stage (M=7.47, SD=15.3); t (11) =2.818, p = 0.0167. These results suggest that students obtained lower average total scores during the recording stage when compared to the average total scores during the interactive stage of the trial. Furthermore a Pearson correlation between the total scores of both stages of 0.0497 indicates a very week correlation between the total scores of both stages.