The city of Gotham published the following statistics on the number over vehicle collisions of the last four years.
- In year 1, 800,000 collisions occurred
- Of these 800,000 - 99% involved combustion engines
- Of these - 73.4% had no injuries and 23.6% had injuries and 3.0% were fatalities
The market penetration of electric vehicles is much smaller than combustion vehicles, but their presence is increasing. No further demographical information such as age, income, cost of vehicle is available.
What statistical tests could be employed to determine if electric vehicles were safer than their combustion equivalents with a degree of statistical significance, given that a motor vehicle collision occurred?
Would a chi-squared test be appropriate? I think that any subtle differences in injured would be masked by the large numbers of uninjured occupants?
Safer could be determined by evaluating the difference in the number of:
- Dead
- Injured
- Injured and Dead