Event History Testing

Abstract
Heterogeneity and nonstationarity typically present in sociological time series have prevented rigorous testing of the axiom of cumulative inertia (McGinnis, 1968). An hypothesis test is presented for event history effects in a collection of short event series controlling for heterogeneity and nonstationarity. The test is applied to the Southeast Wisconsin residential mobility data of Clark et al. (1977, 1979). The results indicate no evidence of any event history effects, and suggest that the residential mobility may be represented by an independent trials process.