Important keywords, points to remember

• Methodology for conducting accident-causation research / model / experimental projcts
• .Case study approach
• Reconstruction
• Probabilistic model
• Statistical inference approach
• Multiple regression model
• Multi dimensional contingency tables
• Computer programs to analyze data from field experiments
• Traffic safety research topics are multidimensional with several levels of many of the variables. Hence a multidimensional contingency approach is preferred
  
• Statistical approach is useful as it can handle complex, interactive phenomena in an objective manner. It allows isolation and study of individual factors
• Statistical approach is preferred over case study approach
• Accident patterns help prediction and computer simulation / modelling
• Conceptual framework for accident related factors is very useful in prediction
• Factors are classified as sequential or operational and simultaneous
• Operational failures are functional failures. Examples are-
• Malfunctioning of perception
• Decision performance in trip planning
• Driving strategy or evasive tactics
• Examples of condition factors are-
• Deficiencies in basic attributes of roads, vehicles and people

• Accident is not a single event. It is a process
• No single factor can explain the accident
• Causal links, each preceded by another
• Each preceding link has lesser relevance to the accident
• By using the above four points, thousands of accidents can be studied by computer analysis (PERCHONOK)
  
• The logic involves division of accident generation process into its component parts and provide an exhaustive list of mutually exclusive items for each of these parts. The resulting series of checklists is called CAUSAL STRUCTURE
• The top level breakdown of causal factors is-
• Human direct causes
• Human condition and state
• Environmental factord
• Vehicular factors

Traffic safety experiment

• A factor F is said to be the cause of accidents if the conditional probability of an accident occurring in the presence of F is greater than the conditional probability of an accident occurring in the absence of F, as determined under condition set E.

• Features of case study approach
• Transient factors
• Dozing
• Lack of attention

• Human factors
• Monocular vision
• Color blindness
• Chronic disease states

• Limitations
• Vehicle handling
• Braking
• Steering
• Human detection
• Perception
• Decision
• Reflexes
• Tinted windshield
• Visual capabilities of driver
• Ambient light levels
• Reflectiveness of different objects

• Features of statistical inference approach
•  Non-accident information
• Accident causation factors
• Stability of factors
• Alcohol
• Loss of consciousness
• Inadequate search
• Detection / perception failures
• Recognition failures
• Decision failure
• Vehicular and environmental failure
• Tire blow-out
• Brake failure
• Complete loss of steering
• The above three cases should be studied as CASE-STUDY and NOT as STATISTICAL INFERENCE method

• Examples of Environmental problems are
• Glare from oncoming headlights
• Smooth road surface due to freezing conditions
• Threaded tires vs Bare tires