Question: Is there a relationship between vehicle damage and collision type in car crashes?
My project will be using data from a dataset that covers car crashes and their variables. I will be using statistical methods to evaluate the relationship between collision type and the damage to a vehicle after an accident. In my dataset there are 39 variables total, and 205539 cases. The data frame collects data from car crashes that have happened in Montgomery county.
Link to the dataset: https://catalog.data.gov/dataset/crash-reporting-drivers-data
Data Analysis In order to address my research question, I should group my prime variables together into one dataset to make comparisons on the severity and extent of vehicle damage. This will also be helpful because inside the damage variability there is range in the severity of vehicle damage by different categories.
##Check structure, and create new data using the variables collision type and vehicle damage to compare
str(dataset)## 'data.frame': 205539 obs. of 39 variables:
## $ Report.Number : chr "MCP3296002G" "MCP276700BF" "MCP32790038" "MCP34000014" ...
## $ Local.Case.Number : chr "240018653" "240012321" "240022955" "240019831" ...
## $ Agency.Name : chr "MONTGOMERY" "MONTGOMERY" "MONTGOMERY" "MONTGOMERY" ...
## $ ACRS.Report.Type : chr "Property Damage Crash" "Property Damage Crash" "Property Damage Crash" "Property Damage Crash" ...
## $ Crash.Date.Time : chr "04/21/2024 06:53:00 PM" "03/14/2024 05:23:00 PM" "05/15/2024 07:30:00 AM" "04/28/2024 05:30:00 PM" ...
## $ Route.Type : chr "County Route" "US (State)" "Maryland (State) Route" "County Route" ...
## $ Road.Name : chr "" "COLUMBIA PIKE (SB/L)" "VEIRS MILL RD (WB/L)" "HARDING DR" ...
## $ Cross.Street.Name : chr "" "" "" "" ...
## $ Off.Road.Description : chr "" "" "" "" ...
## $ Municipality : chr "" "" "" "" ...
## $ Related.Non.Motorist : chr "" "" "" "" ...
## $ Collision.Type : chr "Sideswipe, Opposite Direction" "Sideswipe, Same Direction" "Front to Rear" "Front to Rear" ...
## $ Weather : chr "Clear" "Clear" "Clear" "Clear" ...
## $ Surface.Condition : chr "Dry" "Dry" "" "Dry" ...
## $ Light : chr "Daylight" "Daylight" "Daylight" "Daylight" ...
## $ Traffic.Control : chr "No Controls" "No Controls" "No Controls" "No Controls" ...
## $ Driver.Substance.Abuse : chr "Unknown, Unknown" "Unknown, Unknown" "Unknown, Unknown" "Unknown, Unknown" ...
## $ Non.Motorist.Substance.Abuse : chr "" "" "" "" ...
## $ Person.ID : chr "8F256727-5915-4435-9CE7-5F2D943B8DCA" "BB361322-2BD1-47AE-B6B4-9359BAE743A5" "E04C495C-1DE3-4AC9-B507-A9B2C207056C" "D7E62A5D-C733-490D-AC9C-B5EA089C651E" ...
## $ Driver.At.Fault : chr "No" "No" "No" "No" ...
## $ Injury.Severity : chr "" "No Apparent Injury" "" "" ...
## $ Circumstance : chr "" "" "" "" ...
## $ Driver.Distracted.By : chr "" "" "" "" ...
## $ Drivers.License.State : chr "" "" "" "" ...
## $ Vehicle.ID : chr "FDDB5331-EC22-41B2-9143-124692C70CF6" "7EDA4B0A-1880-4BC7-AB71-DBE62C3A9801" "BBAC2C29-9197-40D8-9B22-9156FA82B6F1" "3ACB0037-C9D5-48C5-B410-BE986F22CA86" ...
## $ Vehicle.Damage.Extent : chr "Disabling" "Vehicle Not at Scene" "Vehicle Not at Scene" "Functional" ...
## $ Vehicle.First.Impact.Location: chr "Eleven O Clock" "Vehicle Not at Scene" "Vehicle Not at Scene" "One O Clock" ...
## $ Vehicle.Body.Type : chr "Passenger Car" "Passenger Car" "Passenger Car" "Passenger Car" ...
## $ Vehicle.Movement : chr "Turning Left" "Overtaking/Passing" "Moving Constant Speed" "Moving Constant Speed" ...
## $ Vehicle.Going.Dir : chr "Northbound" "Southbound" "Northbound" "Southbound" ...
## $ Speed.Limit : int 25 45 45 25 35 35 40 30 5 30 ...
## $ Driverless.Vehicle : chr "No" "No" "No" "No" ...
## $ Parked.Vehicle : chr "No" "No" "No" "No" ...
## $ Vehicle.Year : int 2013 0 2006 2012 2015 0 2015 2006 0 2014 ...
## $ Vehicle.Make : chr "HYUNDAI" "" "CHEVROLET" "NISSAN" ...
## $ Vehicle.Model : chr "SONATA" "" "COBALT" "ALTIMA" ...
## $ Latitude : num 39.1 39 39.1 39 39.1 ...
## $ Longitude : num -77 -77 -77.1 -77 -77.1 ...
## $ Location : chr "(39.10533874, -76.98984545)" "(39.04000338, -76.99722287)" "(39.0730963, -77.1112541)" "(39.02194389, -77.01829961)" ...ds2 <- table(dataset$Collision.Type, dataset$Vehicle.Damage.Extent)
ds2##
## DESTROYED Disabling DISABLING Functional
## Angle 0 4292 0 1385
## ANGLE MEETS LEFT HEAD ON 37 0 367 0
## ANGLE MEETS LEFT TURN 123 0 977 0
## ANGLE MEETS RIGHT TURN 33 0 504 0
## Front to Front 0 1303 0 321
## Front to Rear 0 3186 0 3021
## HEAD ON 541 0 1816 0
## HEAD ON LEFT TURN 1048 0 7881 0
## N/A 10 0 113 0
## OPPOSITE DIR BOTH LEFT TURN 8 0 116 0
## OPPOSITE DIRECTION SIDESWIPE 108 0 862 0
## Other 0 991 0 427
## OTHER 422 0 3442 0
## Rear To Rear 0 55 0 92
## Rear To Side 0 885 0 277
## SAME DIR BOTH LEFT TURN 1 0 83 0
## SAME DIR REAR END 1336 0 16426 0
## SAME DIR REND LEFT TURN 26 0 249 0
## SAME DIR REND RIGHT TURN 19 0 160 0
## SAME DIRECTION LEFT TURN 56 0 1064 0
## SAME DIRECTION RIGHT TURN 58 0 804 0
## SAME DIRECTION SIDESWIPE 209 0 2950 0
## Sideswipe, Opposite Direction 0 448 0 306
## Sideswipe, Same Direction 0 915 0 1300
## Single Vehicle 0 1954 0 485
## SINGLE VEHICLE 1925 0 7840 0
## STRAIGHT MOVEMENT ANGLE 1639 0 16639 0
## Unknown 0 56 0 27
## UNKNOWN 11 0 61 0
##
## FUNCTIONAL N/A No Damage NO DAMAGE OTHER
## Angle 0 0 17 0 0
## ANGLE MEETS LEFT HEAD ON 169 2 0 12 0
## ANGLE MEETS LEFT TURN 589 1 0 17 0
## ANGLE MEETS RIGHT TURN 393 0 0 18 1
## Front to Front 0 0 1 0 0
## Front to Rear 0 0 46 0 0
## HEAD ON 675 0 0 60 0
## HEAD ON LEFT TURN 2697 9 0 74 5
## N/A 81 9 0 123 3
## OPPOSITE DIR BOTH LEFT TURN 90 0 0 11 0
## OPPOSITE DIRECTION SIDESWIPE 632 2 0 106 1
## Other 0 0 17 0 0
## OTHER 4276 83 0 1932 20
## Rear To Rear 0 0 7 0 0
## Rear To Side 0 0 10 0 0
## SAME DIR BOTH LEFT TURN 182 3 0 35 0
## SAME DIR REAR END 18341 91 0 1620 18
## SAME DIR REND LEFT TURN 203 0 0 16 0
## SAME DIR REND RIGHT TURN 215 2 0 31 0
## SAME DIRECTION LEFT TURN 1170 11 0 81 2
## SAME DIRECTION RIGHT TURN 1106 8 0 128 2
## SAME DIRECTION SIDESWIPE 4199 41 0 633 9
## Sideswipe, Opposite Direction 0 0 9 0 0
## Sideswipe, Same Direction 0 0 41 0 0
## Single Vehicle 0 0 205 0 0
## SINGLE VEHICLE 2386 32 0 969 24
## STRAIGHT MOVEMENT ANGLE 7499 18 0 361 17
## Unknown 0 0 15 0 0
## UNKNOWN 90 4 0 17 0
##
## Superficial SUPERFICIAL UNKNOWN
## Angle 961 0 0
## ANGLE MEETS LEFT HEAD ON 0 101 12
## ANGLE MEETS LEFT TURN 0 296 30
## ANGLE MEETS RIGHT TURN 0 234 21
## Front to Front 218 0 0
## Front to Rear 3149 0 0
## HEAD ON 0 597 97
## HEAD ON LEFT TURN 0 1109 103
## N/A 0 178 68
## OPPOSITE DIR BOTH LEFT TURN 0 89 8
## OPPOSITE DIRECTION SIDESWIPE 0 998 174
## Other 480 0 0
## OTHER 0 7180 1675
## Rear To Rear 162 0 0
## Rear To Side 344 0 0
## SAME DIR BOTH LEFT TURN 0 360 55
## SAME DIR REAR END 0 16378 1549
## SAME DIR REND LEFT TURN 0 219 30
## SAME DIR REND RIGHT TURN 0 264 29
## SAME DIRECTION LEFT TURN 0 1192 139
## SAME DIRECTION RIGHT TURN 0 1513 213
## SAME DIRECTION SIDESWIPE 0 6924 1262
## Sideswipe, Opposite Direction 337 0 0
## Sideswipe, Same Direction 2246 0 0
## Single Vehicle 1015 0 0
## SINGLE VEHICLE 0 2357 339
## STRAIGHT MOVEMENT ANGLE 0 3750 419
## Unknown 40 0 0
## UNKNOWN 0 138 397
##
## Vehicle Not at Scene
## Angle 205
## ANGLE MEETS LEFT HEAD ON 0
## ANGLE MEETS LEFT TURN 0
## ANGLE MEETS RIGHT TURN 0
## Front to Front 74
## Front to Rear 614
## HEAD ON 0
## HEAD ON LEFT TURN 0
## N/A 0
## OPPOSITE DIR BOTH LEFT TURN 0
## OPPOSITE DIRECTION SIDESWIPE 0
## Other 153
## OTHER 0
## Rear To Rear 81
## Rear To Side 86
## SAME DIR BOTH LEFT TURN 0
## SAME DIR REAR END 0
## SAME DIR REND LEFT TURN 0
## SAME DIR REND RIGHT TURN 0
## SAME DIRECTION LEFT TURN 0
## SAME DIRECTION RIGHT TURN 0
## SAME DIRECTION SIDESWIPE 0
## Sideswipe, Opposite Direction 120
## Sideswipe, Same Direction 694
## Single Vehicle 239
## SINGLE VEHICLE 0
## STRAIGHT MOVEMENT ANGLE 0
## Unknown 111
## UNKNOWN 0###I can now see the collision type side by side under the categories of vehicle damagesummary(dataset)## Report.Number Local.Case.Number Agency.Name ACRS.Report.Type
## Length:205539 Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
## Crash.Date.Time Route.Type Road.Name Cross.Street.Name
## Length:205539 Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
## Off.Road.Description Municipality Related.Non.Motorist
## Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## Collision.Type Weather Surface.Condition Light
## Length:205539 Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
## Traffic.Control Driver.Substance.Abuse Non.Motorist.Substance.Abuse
## Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## Person.ID Driver.At.Fault Injury.Severity Circumstance
## Length:205539 Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
## Driver.Distracted.By Drivers.License.State Vehicle.ID
## Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## Vehicle.Damage.Extent Vehicle.First.Impact.Location Vehicle.Body.Type
## Length:205539 Length:205539 Length:205539
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## Vehicle.Movement Vehicle.Going.Dir Speed.Limit Driverless.Vehicle
## Length:205539 Length:205539 Min. : 0.00 Length:205539
## Class :character Class :character 1st Qu.:25.00 Class :character
## Mode :character Mode :character Median :35.00 Mode :character
## Mean :32.19
## 3rd Qu.:40.00
## Max. :75.00
## Parked.Vehicle Vehicle.Year Vehicle.Make Vehicle.Model
## Length:205539 Min. : 0 Length:205539 Length:205539
## Class :character 1st Qu.:2006 Class :character Class :character
## Mode :character Median :2012 Mode :character Mode :character
## Mean :1965
## 3rd Qu.:2016
## Max. :9999
## Latitude Longitude Location
## Min. :37.72 Min. :-79.49 Length:205539
## 1st Qu.:39.02 1st Qu.:-77.19 Class :character
## Median :39.07 Median :-77.11 Mode :character
## Mean :39.08 Mean :-77.11
## 3rd Qu.:39.14 3rd Qu.:-77.04
## Max. :39.99 Max. :-75.53##mostly categorical variables in this dataset, besides vehicle year, speed limitdataset <- dataset |>
mutate(dataset$Vehicle.Damage.Extent, recode(Vehicle.Damage.Extent,
"DISABLING" = "Disabling",
"FUNCTIONAL" = "Functional",
"NO DAMAGE" = "No Damage",
"SUPERFICIAL" = "Superficial"))
##combine vehicle damages in the same category into one columnHo: Collision type and vehicle damage are independent to each other
Ha: Collision type and vehicle damage are related/dependent upon each other
##Chi square test of independence between collision type and vehicle damage
chisq.test(ds2)## Warning in chisq.test(ds2): Chi-squared approximation may be incorrect##
## Pearson's Chi-squared test
##
## data: ds2
## X-squared = 288012, df = 336, p-value < 2.2e-16Statistical Analysis The p-value is approxiamte, but still extremely close to 0, proving the alternative hypothesis to be true, leading us to reject the null claim that vehicle collision and vehicle damage are not connected. At the (default) 5% significance level we can conclude that collision type and vehicle damage do share a relationship in car crashes based on the chi test.
Conclusion My chi square test was able to prove that my alternative hypothesis was correct, which was my projection about my project. The findings make sense because the severity, speed, and direction cars crash have a direct impact on the damage of a car. This proves why a fender bender might just end up with scratches on a car, while a t-bone may result in permanent damage to a car’s engine, coolant and oil, etc.
For further analysis or research, I could find a dataset that includes the speed a person was going at the time of a crash, which could further prove how collision type and vehicle damage are related in car crashes, because the faster you go, the more severe the crash may be. This can also be proved through statistical analysis.
##Car crash dataset source https://catalog.data.gov/dataset/crash-reporting-drivers-data ##Data function library for data cleaning https://github.com/rstudio/cheatsheets/blob/main/data-transformation.pdf