Life Expectancy does not Depend on Classical Ecological Variables:Stochastic and Non-stochastic AnalysisAuthor: Umberto Cornelli
The Life Expectancy (LE) in 191 countries was compared on the base of 17 demographic and economic variables. Two approaches were followed: the stochastic and non-stochastic analysis. The first consisted of simple correlation coefficients followed by the, Main Component, Factorial and finally Segmentation Analysis. The second was the Artificial Neuronal Network Analysis using the Auto Contractive Map (Auto-CM). The results of the two approaches were very similar and all excluded at least 10 out the 17 variables. The surface covered by Forests, the Kmq2 of forests, the ratio of the Domestic Gross Profit (GDP) with education, the number of hospital beds, the particulate matter, the population and the population density were not considered determinant. What were emerging as directly correlated with LE were internet, GDPs (GDP/inhab and GDP 2 and GDP 3 related respectively to the advance industry and economy), urban concentration, cars and cellphones. An inverse correlation was found with GDP 1 (related to agriculture, livestock, fishing) and unemployment rate. The last was detected by the ANN only. The conclusion is that LE in the world is far from the variables typically bound to the environment and more linked to the economic variables. One sentence summary: The life expectancy measured with stochastic and non stochastic analysis excludes the correlation with the common ecological variables.