The one on the left is "Chicago-style" seating, with most seats in pairs facing along the length of the car. The one in the middle is "New York-style" seating, with most seats facing sideways. The third is a hybrid.
Transit agencies commonly do surveys that imply that these things are just a matter of taste, as though they'll go with whatever their riders prefer. This question is not just a matter of taste. The left hand image has the most seats but the least capacity. The middle image as the fewest seats but the greatest capacity. Seats with their backs to the wall take up much less space than seats in pairs facing forward or back. And of course, any seat takes up more space than a standee in a crowded car. This is why really crowded subway systems inevitably gravitate toward side-facing seats.
So the question should be not whether you like the the configuration on the left, but whether you like it so much that you don't mind being left behind at rush hour because the train is full.
The survey asks you which configuration you prefer, and which you like better in terms of "personal space." But it doesn't inform the reader that the more forward- and back-facing seats there are, the more people will be left behind on the platform during the peak and the less ridership the system will be able to handle.
Almost all choices are tradeoffs, so when you ask the public their opinion, you need to explain what the real consequences of the options are. (At least that's my firm's approach to public outreach!)