Does reflexive interaction enhance children’s musical improvisation? Computer-child interaction at the beginning stage of music learning

Background Several scholars tackled the issue of children’s musical improvisation from different learning/teaching perspectives and by using different methodologies. We investigated this issue through the reflexive interaction paradigm. This paradigm is based on the idea of letting users manipulate virtual copies of themselves, through specific software called Interactive Reflexive Musical Systems (IRMS). Aims We investigated whether the reflexive interaction using the IRMS influence the children’s skillfulness to improvise, at the beginning stage of musical learning. We used a particular IRMS, the MIROR-Impro, implemented in the MIROR project (European Commission-FP7), which is able to reply to the child playing a keyboard by mirroring (with repetitions and variations) her/his inputs. Method The study was conducted in a primary public school, with 47 children, aged 6 to 7. The experimental design used the convergence procedure, based on three sample groups (one control group - CG - and two experimental groups - EG1 and EG2), allowing to verify if the MIROR-Impro and the reflexive interaction are necessary and sufficient to improve the children's abilities to improvise, in solo and in duet. Independent variables: to play 1) only the keyboard, 2) the keyboard with the MIROR-Impro but with not-reflexive reply, 3) the keyboard with the MIROR-Impro with reflexive reply. Dependent variables: children’s ability to improvise 1) in solo, and 2) in duet. Procedure: each child carried out 5 weekly individual 12 minutes sessions. The CG played the complete package of independent variables (v1+v2+v3); EG1 played the keyboard and the keyboard with the MIROR-Impro with notreflexive reply (v1+v2); EG2 played only the keyboard with the reflexive 002 WWW.PERFORMANCESCIENCE.ORG system (v3). One week after, the children were asked to 1) improvise a musical piece on the keyboard (Solo task), and 2) play the keyboard like in a dialogue, in pair with a friend (Duet task). Activities and test were videorecorded. Results Three independent judges assessed the Solo task improvisations by means of a grid based on the TAI (Test for Ability to Improvise, by Gary McPherson) rating scale. The following assessment criteria were used: Instrumental Fluency, Musical Organization, Creativity, and Musical Quality. The total result shows that the CG (13,56) obtained a lower score compared to both the EG1 (13,82) and EG2 (15,45). Therefore, the total trend reveals that the IRMS and the reflexive interaction are not necessary but sufficient to improve children’s ability to improvise. However, the differences are not statistically significant (p = .35, as from ANOVA). The analyses on each criterion confirm the total trend, in particular as far as the Musical Organization is concerned (CG = 3,13; EG1 = 3,38; EG2 = 4; p = .08). Moreover, the EG2, which played just with the reflexive system, shows a higher score in all criteria. Conclusions The analysis of the Duet tasks is currently ongoing through an original grid for assessing the musical dialogue improvisation. The discussion will focus on: 1) the efficacy of reflexive pedagogy in the field of children’s music performance and technology-enhanced learning; 2) the analysis and assessment of the children’s ability to improvise, and 3) the musical improvisation in computer-child interaction.