An auto-parts store offers a fuel additive that claims to increase a vehicle’s gas mileage. The additive is poured into a vehicle’s gasoline tank after the tank is filled. To measure the claim, two methods to collect data are proposed. Method A: 10 similar police cars owned by a city are selected. After the cars are filled with a fresh tank of gasoline, researchers randomly select 10 of the cars to receive the additive, while the other 10 do not receive the additive. The gas mileage for each car is recorded in miles per gallon.

Method B: 20 customers at an auto-parts store receive free coupons for the additive and are asked to use it with their next fresh tank of gasoline. The customers then report their gas mileage with the fuel additive.

Which method describes an experiment?

Method B is an experiment because every car receives the additive.
Method A is an experiment because the additive is used with a new tank of gasoline.
Both methods are experiments because the additive is used in a new tank of gas in both methods.
Method A is an experiment because random assignment determines which cars receive the additive and which cars do not.