Livestock Studies 2021, Vol 61, Num, 2     (Pages: 046-054)

Examination of OECD Countries for the Presence of Livestock by Non-Metric Multidimensional Scaling

Yıldırım Demir 1 ,Sıddık Keskin 2

1 Department of Statistics, Faculty of Economics and Administrative Sciences, Van Yuzuncu Yil University, Van, Turkey
2 Department of Biostatistics, Faculty of Medicine, Van Yuzuncu Yil University, Van, Turkey
DOI : 10.46897/livestockstudies.610202 - In this study, by Non-metric multidimensional scaling (NMDS) which is one of the multivariate statistical analysis methods, configuration of 36 OECD countries in two dimensional space was examined and similarities/dissimilarities between these countries were determined for the variables regarding with livestock data. As variables, `number of horses, pigs, donkeys, turkeys, mules, geese, goats, sheep, buffaloes, ducks, cattle and chickens` of OECD countries obtained from the website of FAO was used. Euclidean distance was used as distance measures. According to the results of NMDS; USA, Germany, France, England, Canada and Poland had the highest positive effect on livestock. However, Israel, Iceland, Lithuania and Luxembourg were different from other countries with the lowest effects. Similarly, while buffalo was the lowest effective one; pigs, cattle, chicken and turkey were found the highest effective species on the livestock sector. As a result, it was suggested that NMDS can be used as an effective method in the analysis of multivariate data in agriculture and livestock with simple graphical representation and interpretation of the results. Keywords : Distance Stress Dimension reduction Configuration Similarity