Sensitivity analysis in the context of regional safety modeling: Identifying and assessing the modifiable areal unit problem

Support vector machine in crash prediction at the level of traffic analysis zones: Assessing the spatial proximity effects

“…Regarding the socio-economic factors, median household income is negative, suggesting that TAZs with a lower median household income are relatively associated with a worse safety situation. The association between median household income and crash risk has been well recognized in previous road safety analysis (Noland and Quddus, 2004;Xu et al, 2014). In addition, area of housing units is found to have positive effects on crash rate, implying more residents in an area have more activities that could results in more traffic crashes.…”

“…Previous studies (i.e., Quddus 2008;Siddiqui et al, 2012;Xu et al, 2014;Zeng and Huang, 2014a) found that traffic crashes exhibit extensive spatial dependency across neighboring zones. Research commonly seeks to address the issue of unmeasured spatial correlations using spatial econometric methods among the neighboring spatial units for two reasons: (a) the collection of crash data observations associated with the spatial units does not accurately reflect the nature of the underlying process that generates the sample data, which might induce measurement errors (Anselin, 2001); (b) the spatial dimensions of socio-demographic, economic or regional activities may truly represent an important aspect in model development and may help to improve the accuracy and robustness of crash prediction and avoid underestimation of standard errors for model parameters.…”

Modeling crash spatial heterogeneity: Random parameter versus geographically weighting

“…Various area-wide characteristics were considered, including road characteristics such as intersections density (e.g., , road length with different speed limits (e.g., Abdel-Aty et al, 2011;Siddiqui et al, 2012), road length with different functional classifications (e.g., Quddus, 2008;Hadayeghi et al, 2010;, junctions and roundabouts (e.g., Noland and Quddus, 2004;Quddus, 2008); traffic patterns in terms of volume and speed (Noland and Quddus, 2005;Quddus, 2008;Hadayeghi et al, 2010); trip generation and distribution (Abdel-Aty et al, 2011); environmental and weather conditions (Aguero-Valverde and Jovanis, 2006); land use (e.g., Hadayeghi et al, 2010;Siddiqui et al, 2012;Pulugurha et al, 2013); and socioeconomic factors such as population density (e.g., Hadayeghi et al, 2006;Siddiqui et al, 2012), age cohorts (e.g., Aguero-Valverde and Jovanis, 2006;Quddus, 2008;Hadayeghi et al, 2010;, household incomes (e.g., Siddiqui et al, 2012;Xu et al, 2014) and employment (Quddus, 2008;Hadayeghi et al, 2010;Xu et al, 2014).…”

Spatiotemporal and random parameter panel data models of traffic crash fatalities in Vietnam

“…block groups (Dumbaugh and Rae, 2009), grid structure (Kim et al, 2006), census tracts (Wier et al, 2009), wards (Noland and Quddus, 2004), cantons (Aguero-Valverde, 2013), counties (Traynor, 2008;Huang et al, 2010), provinces (Erdogan, 2009;Tolón-Becerra et al, 2012) (Moeinaddini et al, 2014;Coruh et al, 2015), multiple provinces (Torre et al, 2007), states (Noland, 2003), countries (Kumara and Chin, 2004), and traffic analysis zones (TAZs) Hadayeghi et al, 2003;Lovegrove and Sayed, 2006;Pirdavani et al, 2012;Wang et al, 2013). Effects of spatial units on modelling results have been discussed in few studies Lee et al, 2014b;Xu et al, 2014).…”

“…spatial dependence or correlation and spatial heterogeneity (Anselin, 1988), have been considered in macro-level safety analyses. For example, Bayesian spatial approaches have been used to account for possible spatial correlation between areas (Aguero-Valverde and Jovanis, 2006;Quddus, 2008;Siddiqui et al, 2012;Wang et al, 2013;Xu et al, 2014;Dong et al, 2015;Lee et al, 2015;Song et al, 2015;Siddiqui and Watkins, 2016). To consider spatial heterogeneity, previous macro-level safety studies have adopted the geographically weighted regression (GWR) models (Hadayeghi et al, 2003;Erdogan, 2009;Hadayeghi et al, 2010;Li et al, 2013) and random parameter models (Coruh et al, 2015;Xu and Huang, 2015).…”