A latent class analysis of single-vehicle motorcycle crash severity outcomes

Hierarchical Bayesian random intercept model-based cross-level interaction decomposition for truck driver injury severity investigations

“…Finite-mixture (latent-class) models are presented with different model structures and have been gaining their popularity in traffic safety analysis (Eluru et al, 2012;Lemp et al, 2011;Shaheed and Gkritza, 2014;Xie et al, 2012;Xiong and Mannering, 2013;Zou et al, 2014Zou et al, , 2013. For example, Shaheed and Gkritza (2014) utilized a multinomial logit model with two latent crash data classes to investigate crash severities in single-vehicle motorcycle crashes. Zou et al (2013) advocated that weight parameter configuration is preferred in finite mixture negative binomial models to better assess heterogeneity effects in crash data analysis, and they further developed different functional forms for weight parameter estimation (Zou et al, 2014).…”

A comparison of the mixed logit and latent class methods for crash severity analysis

“…However, few studies have used a latent class approach for analyzing crash (or driver injury) severities (Chu, 2014;Eluru et al, 2012;Shaheed and Gkritza, 2014;Xie et al, 2012;Xiong and Mannering, 2013;Yasmin et al, 2014aYasmin et al, , 2014b. Datasetspecific evidence in the literature suggest stronger statistical support for latent class logit models for exploring crash severity compared to conventional multinomial logit models (Xie et al, 2012), ordered logit models (Chu, 2014), and generalized ordered logit models (Yasmin et al, 2014a(Yasmin et al, , 2014b.…”

Effects of Human-Centered Factors on Crash Injury Severities

“…Ignoring the moderating effect of unobserved variables can lead to biased estimates and incorrect inferences if inappropriate methods are used [66,67]. Limiting the impact of a variable to its statistical significance in a model can mean eliminating some otherwise risky factors.…”

“…However, these model forms can restrict the way variables influence outcome probabilities, possibly leading to incorrect inferences [37,70]. Compared to the traditional ordered probability models, multinomial logit (MNL) models have a flexible structure which allows each severity outcome to have a different function for capturing the probabilities of injury severities [66,71,72]. Notwithstanding this, the MNL model is deficient in its application as it is susceptible to correlation of unobserved effects from one crash severity level to the next.…”

“…To select the model that best fits the data, likelihood ratio tests may be performed to compare models with different number of classes [75], or based on the Bayesian Information Criterion (BIC) computed for the two models [83][84][85][86]. Recent studies of crash injury severities have used the BIC measure to determine the number of classes [63,64,66]. The BIC for a given empirical model is equal to…”