The participant votes between a pair of answers (A vs. B), without the option to vote for the same answer repeatedly. Showing the voter several different responses prevents them from unduly influencing results.