As the IIHS and other safety agencies around the world have begun testing electric and other alternative-fuel cars, it seems worth asking whether our readers see these tests as more or less important than those for conventional cars.

It isn't that the cars are more or less safe than the best gas-powered cars. Just last week, the IIHS released ratings for the Audi E-tron quattro—an electric SUV with a 95-kilowatt-hour lithium-ion battery pack—and the Hyundai Nexo—a fuel-cell powered SUV carrying 6.3 kilograms of hydrogen compressed to 10,000 psi. Both earned the institute's strenuous Top Safety Pick+ award.

Last month, the U.S. government's NHTSA and the European Euro NCAP safety ratings agency gave top marks to the Tesla Model 3 in their latest testing. (The IIHS has not yet finished testing the Model 3.) 

Hyundai Nexo crash test

Hyundai Nexo crash test

Our question isn't whether these ratings should be figured differently than for standard gas cars (they aren't). It's whether buyers of such machines have greater or fewer concerns about their safety than they do about conventional gas cars.

Gas cars have been around for well over 100 years, and consumers well understand their safety risks and how to minimize them. These new cars, whether they run on lithium-ion batteries or compressed hydrogen, have shorter track records. News reports proliferate whenever an electric car catches fire—in part in an effort to understand these new technologies better—even though more than 100,000 gas cars also catch fire every year. Consumers understand and accept those risks. The same concerns may apply to internal combustion cars that run on natural gas, for example.

Our Twitter poll question this week asks: "Are safety ratings more important for EVs vs gas?"

The range of options we've included are: electric cars, gas cars, other (which could include hydrogen fuel cells), and all are equal.

Let us know what you think by clicking over to our Twitter poll. And remember that our Twitter polls are not scientific because we generally get too few respondents and our respondents are self-selected anyway.