MEASURING THE RELATIONSHIP BETWEEN HUMAN DEVELOPMENT INDEX (HDI) AND HAPPY PLANET INDEX (HPI) WITH CANONICAL CORRELATION ANALYSIS


Yücel L.

7. INTERNATIONAL ANTALYA Antalya, Turkiye 11-13 May 2024 SCIENTIFIC RESEARCH AND INNOVATIVE STUDIES CONGRESS, Antalya, Turkey, 11 - 13 May 2024, pp.18

  • Publication Type: Conference Paper / Summary Text
  • City: Antalya
  • Country: Turkey
  • Page Numbers: pp.18
  • Istanbul University Affiliated: Yes

Abstract

Introduction and Purpose: In this study, the relationship between the Human Development Index (HDI) and the Happy Planet Index (HPI), which were created based on the idea that people's happiness, welfare and development cannot be explained only by economic growth, is tried to be measured by Canonical Correlation Analysis. Since both indices are like a data set rather than a variable with their structure consisting of 3 sub-components each, it is aimed to show how the relationship between them is misleading when measured by Classical Pearson Correlation Analysis. This is because the Classical Pearson Correlation Analysis considers the variables in pairs and cannot make a holistic inference. Canonical Correlation Analysis, on the other hand, considers the variables as a whole with their sub-components and measures the relationship through the canonical variables it creates. Materials and Methods: The canonical relationships between HDI and HPI were measured for 34 OECD countries. Korea, Costa Rica, Luxembourg and Turkey were excluded as some data were not available for these 4 countries. The data of the study belongs to 2020, which is the most recent common cluster for HDI and HPI. SPSS Syntax module was used in the application. Findings: Three pairs of canonical variables were calculated between HDI and HPI variable sets, but only two of them were found to be statistically significant. The first canonical correlation value is 0.976 (Wilks L=0.017) and the second canonical correlation value is 0.799 (Wilks L=0.357). The first canonical variable of HDI is mainly composed of “life expectancy at birth index”, secondly “GDP index” and thirdly “education index”. The second canonical variable of HDI is mainly composed of “education index” (in negative direction), followed by “GDP index” (in negative direction). The first canonical variable of the HPI is predominantly composed of “life expectancy at birth” and secondly “welfare level”, the effect of “ecological footprint” is not significant. The second canonical variable of the HPI is mainly composed of “ecological footprint” (in negative direction) and secondly “welfare level” (in negative direction). “Life expectancy at birth” was not significant in this canonical variable. Results: This study has shown that the relationship between HDI and HPI can be as high as 0.976 when considered holistically with Canonical Correlation Analysis. The Classic Pearson Correlation value between HDI and HPI is 0.16. This shows that if the linear relationship between only two variables is not to be measured directly, as in this study, and if the relationships between variables with sub-components are to be measured, Canonical Correlation Analysis is a more appropriate way than Classical Pearson Correlation Analysis.