MEFEMCO: Methodological Foundations of Embodied Music Cognition

Project Description:

The MEFEMCO project aims at de development of an empirical methodology for embodied music cognition (see Leman, 2007).

Methodology

The development of new technologies at IPEM will make it possible to measure the gestural behavior of humans as a response to musical stimuli. The resulting experimental designs are complex and multidimensional and give rise to large datasets. Multidimensionality is the consequence of different sources of variation that may have an influence on human musical gesture. In order to analyze the combination of human factors, socio cultural context, subjective experience, musical parameters and physical energy, specific Multivariate Statistical Paths need to be developed. A broad spectrum of traditional statistical techniques such as advanced Time-Series Analysis, Multivariate Non-Parametric Analysis, Multivariate Analysis of Variance (GLM) and Multiple Regression Analysis will be used. Furthermore, new statistical methods, such as Dynamic Time Warping, Chaos Analysis, Cladogram Analysis and Analysis of Entropy, will be explored as possible other valid methods for complex data analysis.

Research Objectives

The objectives of the combination of standard multivariate analysis and new analytical tools are:

  • setup of complex experimental designs
  • investigation and analysis of the multidimensional gestural space
  • development of statistical paths as practical guides to make complex statistical analysis accessible for other researchers at IPEM

 

Promotors: Prof. Dr. M. Leman (IPEM : Dept. of Musicology - Ghent University)
Dr. Dirk Moelants
Collaborators: Frank Desmet
Researchers: Frank Desmet
Financial Support:

Ghent University - Fund for Scientific Research (FWO) - Flanders (Belgium)