Background: Official statistics in Taiwan have shown that, following hook-turn violation, motorcyclist’s red-light violation (RLV) is one of the leading traffic violations. RLV by electric motorcycles/bicycles is particularly serious when there has been an increased use of electric motorcycles and bicycles in recent years. Bicyclists’ red-light violations (RLVs) tend not to cause accidents although RLV is a frequent and typical bicyclist’s behaviour. Aims: The current research aims to investigate the contributory factors to motorcyclists’ and bicycles’ RLVs. Method: The present study will employ an observational survey using video/speed cameras to capture the outcome variable of interest (i.e., motorcyclists’ and bicyclists’ crossing behaviours), as well as other independent variables (including rider attributes, temporal factors, vehicle attributes, and roadway characteristics). The duration of the project is two years: in first year, the current research will firstly observe motorcyclists’ and bicyclists’ crossing behaviours (when facing red lights) at the selected rural (New Taipei City) intersections, and in second year, at urban (Taipei City) intersections. All two wheeled users (motorcycles, bicycles, electric motorcycles/bicycles) facing the red lights (excluding amber light) at the selected intersections will be the subjects, providing a rich source of the observations that facilitates the statistical modelling of the determinants of crossing behaviours. The subjects’ crossing behaviours will be classified into three distinct manners (i.e., risk-taking, opportunistic, and law-obeying), as certain groups of motorcycle/bicycle users may have more tendencies to commit one of the three crossing manners. For instance, users with slower power output (i.e., electric motorcycle/bicycles) may be more likely to be an opportunistic violator, while standard users (e.g., motorcycles) would be more prone to be a risk-taker. The current research will conduct advanced statistical models to account for the unobserved heterogeneity that may exist in an observational study. Model estimations results of typical logit models will be compared against those of advanced statistical models (e.g., random parameters logit model, latent class model). Expected outcomes: A better understanding of the influential factors will facilitate the identification of suitable policies/strategies that could further curb the RLV problems that are specific to motorcyclists, and in turn reduce accident risks at junctions. Likely research outcomes include that several rider groups (e.g., mopeds, heavy-engine motorcycle riders, young riders), certain time intervals (off-peak hours) will be associated with risk-taking behaviours, while there would be a prevalence of opportunistic manner during peak hours, at small junctions, and when the riders are travelling on electric motorcycles/bicycles.
|Effective start/end date||8/1/15 → 10/31/16|
- Crossing behaviour
- Red-light violation
- Electric motorcycle/bicycle
- Random parameters logit model
- Latent class model