“…Although it could be greatly desirable and beneficial to conduct the safety analysis on the basis of traffic crash data, there are three critical issues associated with the existing crash database in China: (1) the crash data are heavily underreported, e.g., the property damage only data are typically not included and investigated by the police (Cheng, 2004;Fei et al, 2006); (2) the officially published crash counts are seriously biased due to the political concerns (Fei et al, 2006); and (3) the design layout of LWA at the signalized intersections is relatively new as opposed to the conventional intersections (and currently available only in the capital cities of China) and thus the number of crashes occurring at intersections as such is limited. Also, the historical research works (Gettman and Head, 2003;Oh et al, 2010;Autey et al, 2012;Shahdah et al, 2014) have indicated that the development of safety evaluation based upon the traffic conflict technique can well reflect the safety situation. Comparatively, traffic conflict data are easier to collect and measure as opposed to the crash events (due to the crash characteristics of randomness and rarity) at the intersections with LWA.…”