We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Estimation Issues Associated with Time-Series--Cross-Section Analysis in Criminology.
- Authors
Worrall, John L.; Pratt, Travis C.
- Abstract
In this paper we offer a relatively comprehensive introduction to estimation issues associated with time-series—cross-section analysis in criminology. We divide the estimation issues into two categories: (1) those that have received a fair amount of attention in the literature and (2) those that have not. Issues that have received attention are heterogeneity, autocorrelation, panel heteroskedasticity, nonstationarity, and unit-specific trends. Issues that have not received much attention are spatial autocorrelation and contemporaneous correlation. Using county-level data from the state of California, focusing in particular on the crimes of assault, robbery, and burglary, we control for the first set of estimation problems, then we explore the effects of the latter set. We conclude that contemporaneous correlation deserves more attention than spatial autocorrelation. Also, we found that assault is more sensitive to the estimation issues raised than either robbery or burglary.
- Subjects
ESTIMATION theory; TIME series analysis; CRIMINOLOGY; HETEROGENEITY; AUTOCORRELATION (Statistics); STATISTICAL correlation
- Publication
Western Criminology Review, 2004, Vol 5, Issue 1, p35
- ISSN
1096-4886
- Publication type
Article