RT Article
T1 Estimation issues associated with time-series: cross-section analysis in criminology
JF Western criminology review
VO 5
IS 1
SP 35
OP 49
A1 Worrall, John L.
A1 Pratt, Travis C.
A2 Pratt, Travis C.
LA English
YR 2004
UL https://krimdok.uni-tuebingen.de/Record/1639839496
AB 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.
NO Literaturverzeichnis: Seite 46-48
K1 Methodenfragen
K1 Kriminologische Forschung
K1 Zeitreihen
K1 time-series--cross-section data
K1 Heterogeneity
K1 autocorrelation
K1 panel heteroskedasticity
K1 Nonstationarity