Traffic Accidents

Only available on StudyMode
  • Download(s) : 1618
  • Published : March 9, 2011
Open Document
Text Preview
Vehicle Accident Analysis
BUDT 733, Prof. Schmueli
Amelie Brandenberg Jason Dell Donna Donella Rama Reddy

Outline
Issue - Vehicular Accident Injuries Project Objective Data source and variables Research questions Methods of analysis – – – –

Exploratory analysis Descriptive statistics Logistic regression Discriminant analysis

Results Recommendations

Factors in Vehicular Accidents
Physical environment Person – driver, passenger Vehicle related Other

Data Source
The Bureau of Transportation Statistics (part of the U.S. Dept. of Transportation) gathers data on the estimated 6.4 million vehicle accidents reported each year. Data are accessible at http://www.transtat.bts.gov. The most recent sample of 55,000 records relates to 2001 vehicle accidents. From this data, we selected a random sample of 10,000 records for this project.

Objective
Objective
– Evaluate the role of physical environment in

vehicle accidents that result in injuries, specifically
Profile variables of physical environment that increase the probability of injuries in vehicular accidents Predict the probability of injuries in accidents based on the selected variables of physical environment

Hypothesis
– Variables of physical environment have major influence on the

probability of injury in vehicular accidents

Clients for this Study
Federal Agencies and Dept. of Transportation State and County Governments Automobile Industry Insurance Industry – Health, Auto Businesses in road safety Commuters

Data - List of Variables
Hour of accident - 0-24 hrs Alcohol Road alignment Manner of collision Traffic lanes - one /two/more Traffic conditions Road surface Population density Weekday Interstate Highway Relative to Roadway Work zone Relation to Junction Number of travel lanes Region - NE/MW/south/west Month Road surface conditions Traffic Flow Pedestrian/cyclist Light condition Speed limit Weather

Sample Data
CASENUM 110215646 110215716 110215725 110215728 110215741 110215744 110215764 110215767 110215782 110215791 110215808 110215823 110215824 110215880 110215887 110215889 110215896 110215925 110216025 110216026 110216061 110216072 110216073 110216134 STRATUM WRK_ZONE 1 3 1 2 1 4 1 2 1 1 4 1 4 1 1 4 4 1 1 4 1 4 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 WKDY_I 7 6 3 2 1 3 1 6 5 7 1 1 7 3 7 7 3 3 5 1 2 7 3 4 EVENT1_I 25 25 25 25 25 25 25 45 25 37 45 37 25 25 37 45 25 35 38 25 25 24 38 25 NO_INJ_I INJURY_CR ASH 3 1 1 3 0 0 0 4 0 0 3 1 1 1 0 0 0 0 1 0 0 4 0 0 0 1 1 0 0 0 1 0 0 1 1 1 1 0 0 0 0 1 0 0 1 0 0 0 M AXSEV_I 3 2 1 0 0 0 2 0 0 1 1 1 1 0 0 0 0 1 0 0 1 0 0 0 INT_HWY 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Research Questions
Data are explored with focus on following questions
- What time of day the accidents result in most injuries? - Are road alignment and profile factors in accident injuries? - To what extent is speed limit a factor in injuries? - Do seasonal factors – spring/summer/winter explain injuries? - Whether manner of collision is a factor in the injury? - Whether region affects the probability of injury? - Do number of vehicles involved explain the injury? - Do accidents on interstate highways result in more injuries

Methodology
Started with data organization, exploratory analysis and descriptive statistics Selected variables for modeling the relationships. Logistic regression to predict the odds of injury Discriminant analysis to predict injury/non-injury group membership Evaluation of model results Recommendations from the analysis

Exploratory Analysis
Accidents involving injuries occur most frequently in the afternoon hours. 450 Sum of INJURY_CRASH 400

350

300

250

200

150

100

50

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

HOUR_I

Injuries by Week and Month
MONTH 1 2 3 4 5 6 7 8 9 10 11 12 Total Percent 1 41 31 32 52 41 38 48 60 43 38 45 35 504 10.4 2 47 37 51 56 51 61 72 50 45 82 46 56 654 13.5 3 72 59 61 52 82 61 89 57...
tracking img