Methods and Data

Abstract

It will be my task to evaluate the theoretical relevance of the above presented theories in an empirical, cross-national investigation of world development from 1960 onwards. Modernization, dependence/world system approaches and socio-liberal traditions will be submitted to politometric testing on the basis of cross-national data. The standard statistical technique of ordinary least square multiple regression, as installed in the standard statistical social science package SPSS IX is applied to test for the quantitative relationship between various explanatory variables X₁ to X _n and the explained variable Y, with ‘a’ being the intercept and b₁ to b_n the unstandardized regression coefficients:

$$Y = a + {b_1} * {X_1} + {b_2} * {X_2} + {b_3} * {X_3} + {b_n} * {X_n} + e]$$

((6.1))

In my equation, e is the error term. When we z-standardize my equation, beta₁ to beta_n are the standardized regression coefficients. Significance is the measure, according to which a relationship falls within the boundaries of what must be expected by simple random, or, due to its extraordinary strength, is something not to be explained by random: i.e., it is significant. The level of significance can be interpreted in terms of the statistical probability of a relationship given the numbers of cases and explanatory independent variables, and is designated by p, the probability. Adjusted for the number of variables and cases (‘degrees of freedom’) we arrive also at a measure, which compares the ‘band spread’ of the explained variable, ^Y, (variance), as predicted by my equation, compared to Y, observed in the real world.