China
Africa
Explore chapters and articles related to this topic
Concert Hall Acoustics
Published in Nick Zacharov, Sensory Evaluation of Sound, 2018
The individual profiles were alone of interesting, however the main objective was to reveal the common sensory space amongst the assessors. To this end we have employed hierarchical multiple factor analysis (HMFA, Abdi et al. (2013)) together with agglomerative hierarchical clustering (Lucas, 2008) of the attributes. Multiple factor analysis (MFA) is used in the sense of Escofier and Pagès (1994) using the implementation in FactoMineR package in R (Lê et al., 2008). It is basically as an extension of PCA to multi-block and multi-table data sets. HMFA can effectively handle data sets where the same set of stimuli (samples) are being evaluated by different sets of variables (attributes) in an overall hierarchical structure. The results of such an analysis are a latent space, comprising of multiple dimensions similar in nature to that of a PCA. One benefit of this approach is that instrumental variables, such as the objective ISO 3382-1 (2009) acoustic measures can also be incorporated into the same latent space in order to investigate their relationships with the descriptive data1.
Constructing and measuring domain-specific emotions for affective design: a descriptive approach to deal with individual differences
Published in Ergonomics, 2020
Mingcai Hu, Fu Guo, Vincent G. Duffy, Zenggen Ren, Peng Yue
Unlike classic emotion data set that is quantified by unified EWs, the individualised emotion data grids obtained from the RGI/RATA cannot be analysed by the traditional multivariate tools such as factor analysis. Formally, multiple sets of variables have been measured on the same set of observations (i.e. domain elements for here). The number and/or nature of the variables used to describe the observations can vary from one set of variables (i.e. one interviewee’s own EWs for here) to the other. This problem of mixed data can be appropriately addressed by a recent technique called Multiple Factor Analysis (MFA) that originates from the work of Escofier and Pagès (1988). MFA has been recently used in various domains which encounter the problem of mixed data as well, such as sensory and consumer science research (Naes et al. 2017) and neuroimaging (Buchsbaum et al. 2012) to name but a few.