Naslov
Multivariate Statistical Methods in Researching Biocultural Diversity

Autori
Sindik, Joško ; Šarac, Jelena

Izvornik
Environment and ecology research (2331-625X) 4 (2016), 1; 21-29

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
Association; Biodiversity; Biocultural and Cultural Diversity; Nonlinear Method

Sažetak
Nonlinear multivariate statistical methods have proven to be useful tools in research issues dealing with biocultural diversity. Namely, these methods have less restrictions in their use, as compared with compatible linear methods. This research is the example of using some of these methods. The three indices of biocultural diversity by the variables of biological and cultural diversity have been predicted, based on population size (POP), areal size (AREA) and overall biological and cultural richness (RICH). Then, we have determined: clusters in which different countries can be grouped based on biocultural diversity indices (POP, AREA, RICH), the latent dimensions of the biocultural diversity in the space of biocultural diversity indices (POP, AREA, RICH) and finally, the association between the indices of biodiversity and cultural diversity (POP, AREA, RICH). General conclusion is that nonlinear multivariate methods (together with cluster analysis), in spite of their robustness, can provide useful information to the researchers, on the issues they are interested in (in this example, about biocultural diversity). Of course, these methods are more convenient than linear methods only in the context of absence of clear linear relationships between the variables of interest, while the overfitting problem couldn't be avoided in both cases.

Izvorni jezik
Engleski

Znanstvena područja
Matematika, Biologija, Etnologija i antropologija

POVEZANOST RADA