LOCALIZED AUTOCORRELATION DIAGNOSTIC STATISTIC (LADS) FOR SOCIOLOGICAL MODELS - TIME-SERIES, NETWORK, AND GEOGRAPHIC DATA SETS

Regression models in sociology, because they are often based on datase ts with a surfeit of variables and an underlying connectivity pattern, permit the we of unique diagnostic techniques. This article elaborate s on the localized autocorrelation diagnostic statistic, LADS, which d etermines the probability that in a model with N cares, a connected se t of size C or more among the E most extreme, same-signed residuals oc curred by chance. LADS can suggest variables to be included in a model and can be applied to time-series, geographic group(i.e., cliques, bl ocks, clusters, and different values on a nominal variable), and netwo rk data Exact formular for LADS for time-series and grouped data, as w ell as principles for the robustness of LADS under global autocorrelat ion, are introduced and a general algorithm for all data sets of conne cted cases is presented Examples demonstrate how LADS can suggest new variables and improve the overall fit of models.