The Economist has an article outlining an experimental result that I found surprising: using a hands-free for chatting in the car significantly impairs driving ability.

What makes this surprising to me is that an earlier study had shown that chatting to a passenger in the car does not impact driving ability.

Why would chatting to someone in person be OK but chatting on a hands-free have a big impact on your reactions?

I think that this may have something to do with context. A passenger understands a break in conversation when the driver needs to concentrate on something. More experiments are needed…

Another experiment showed that listening to someone talking has no impact on driving ability – even when subjects knew they would later be tested. So the problem clearly arises when drivers must think of responses.

PS. The researchers found that having a passenger was safer because the passenger would comment on things happening on the road.