Social indicators: A statistician's view

TLDR: The problem areas of social indicator research are considered, what social variables should be considered as conceivable variables related to quality of life, what data should be collected taking in account the difficulty of not being able to directly measure variables of interest, and how to collect data guarding against multicollinearity.

Abstract: In the following we consider the problem areas of social indicator research which are of concern to the statistician and in which he can prove helpful. Among these are the purposes of social indicators, what social variables should be considered as conceivable variables related to quality of life, what data should be collected taking in account the difficulty of not being able to directly measure variables of interest, how does one collect the data (which is usually in the form of a time series) guarding against multicollinearity, and how should the collected data be handled and analyzed. We discuss why in social indicator research the secular trends, cyclical movements, seasonal variations and irregular fluctuations must be taken into account. Techniques are discussed for relating lead indicators in one time period to coincident indicators in another period. Finally we present a select bibliography in canonical correlation, forecasting, indicators and index numbers, path analysis, regression analysis, simulation techniques, time series analysis and other areas useful in analyzing social indicator data.