This chapter begins with a brief introduction to time-series analysis, a method essential to the forms of functional analysis advocated in this book. Included in this introduction are computer simulations of timeseries processes, presented solely for their didactic and heuristic value. Several published works are cited in order to illustrate various degrees of success and failure in studies of social dynamics, and it is concluded that powerful forms of functional analysis will be forthcoming when survivorship variables are incorporated into adaptation hypotheses tested by means of modern methods of time-series analysis.

A primer on time-series analysis

Suppose that in a multivariate causal model we have n variables of which k are endogenous, meaning that they are determined by other variables in the model. In addition there are n – k exogenous variables, defined as those taken as “givens” of the model because they are not explained with reference to other variables included in the model. The several explanatory variables found in any equation may be correlated with each other, but as long as they are not highly correlated we usually do not have to worry about this feature.

In designing complex causal models, it is important that we make the equations representing them amenable to solution. Multivariate causal models, often consisting of several interrelated equations, cannot be solved unless they are “identifiable.”