About two weeks ago, I got into a debate with a friend about Adam Driver. The question at hand: “Is Adam Driver a millennial sex symbol?” What do I mean by that? The answer depends. You could focus on the “millennial” part, and argue that Adam Driver’s status as a GQ cover model is emblematic of the absurdism of the world that millennials have grown up in. You can instead focus on the “sex symbol” part, to which the answer is rarely an enthusiastic yes. After a long and heated debate, I did what anyone would do: I designed a Qualtrics survey pitting three options against each other, randomizing them, and then asking about all three options put together.
The results? “Why? Uh, I mean I guess” wins against both other options in pairwise matchings. This answer is a Condorcet winner.
However, when all options are compared against each other, this answer finishes last (and by a lot).
Why does this happen? And what does this mean? First, this survey is structured according to the assumptions of the median voter theorem: the spectrum ranges from “No” to “Why” to “Of Course,” with “Why” in the middle. The final survey shows which options are people's’ “first” choices. Obviously, for those who prefer “Of Course,” but faced with the choice between “No” and “Why” will choose “Why.” Similarly, those who prefer “No,” but faced with the choice between “Of Course” and “Why” will also choose “Why.”
The second reason points to the structuring of the agenda. When the only options are “No” and “Of Course,” “No” is a clear (albeit slight) winner. However, when all three options are included, no is not a winner. It places second in three way voting (with “Why” playing the spoiler) and still loses head to head against “Why.”
The takeaway from this is that the Condorcet winner is not always the best choice. It seems to be the most “agreeable” choice, but if it’s included just to be a spoiler, it may detract from true majority rule decision making.
No comments:
Post a Comment