Relative Distribution Methods In The Social Sciences

Understanding Distributional Differences in Social Science Research The study of differences among groups or changes over time is a common focus in social science research. However, traditional statistical methods often fail to capture the underlying social theory's implications for distribution properties. This can be seen in debates around growing inequality, racial differences in test scores, socio-economic correlates of birth outcomes, and the impact of smoking on survival and health. These complex distributional differences involve more than just mean-shifts and changes in variance; they also encompass subtle comparisons of changes in the upper and lower tails of distributions. Traditional data analysis methods often leave this information untapped. This book presents methods for full comparative distributional analysis, based on the relative distribution, a nonparametric complete summary of the required information for scale-invariant comparisons between two distributions. The relative distribution provides a general integrated framework for analysis, offering a graphical component to simplify exploratory data analysis and display, a statistically valid basis for developing hypothesis-driven summary measures, and potential for decomposition that enables examination of complex hypotheses regarding distributional changes within and between groups.