RT Article
T1 Utilizing alternate models for analyzing count outcomes
JF Crime & delinquency
VO 63
IS 1
SP 61
OP 76
A1 Rydberg, Jason
A2 Carkin, Danielle Marie
LA English
YR 2017
UL https://krimdok.uni-tuebingen.de/Record/1577205472
AB Although ordinary least squares (OLS) regression was once a common tool for modeling discrete count outcomes in criminology and criminal justice, the past several decades have seen an increasing reliance on regression techniques specifically designed for such purposes. Utilizing a practical example from the 1958 Philadelphia Birth Cohort, this article describes and compares various estimation strategies for modeling such outcome variables, including a discussion of the inappropriateness of OLS for such purposes and specific features of discrete count distributions that complicate statistical inference—overdispersion, non-independence, and excess zeros. Practical advice for selecting an appropriate modeling strategy is offered.
K1 Quantitative
K1 Count regression models
K1 Zero-inflated models
K1 Hurdle models
DO 10.1177/0011128716678848