TIL about the "unmatched count technique" in statistics / polling.
Basically, it's a way to ask people embarrassing questions that doesn't require them to give a direct answer, and thus makes them more likely to answer honestly.
The way it works, is to have a list of like 10 totally questions or so, and then give that to one group of people. Instead of answering yes/no for each question individually, they answer with a single number saying how many of those statements are true.
I found out about this from a paper that claims that the real number of atheists is significantly higher than what any of the direct polling shows: osf.io/preprints/psyarxiv/edzda (although I'm not gonna get into that debate here.)
> Rather surprisingly,and in contrast to previous unmatched count validation(Coffman et al., How many atheists are there? 132016), we observed a reliable difference between these two conditions, with people indicating more statements true of them in the mathematical impossibility condition (M= 3.62, SD= 1.16) than in the baseline condition (M= 3.27, SD= 1.15), mean difference = .340 [.198, .477]. This result is, frankly, bizarre and we are hesitant to speculate a great deal about its causes.
Where the mathematical impossibility condition is "I do not believe that 2 + 2 is less than 13.".
@El'endiaStarman might be trolling... might also be the negative? As in, asked how many statements are true, and they are presented with a true math statement but are expected to answer true if the math is wrong.
The prompt was "How many of the statements below are true of you?" with a list of six innocuous statements and that one (or "I do not believe in God.").
My university has something called "dining dollars" AKA fake money that you can use to buy food.
The upside is that paying with dining dollars is tax-free and so you save 6-7% on every purchase, and you get some free meals with the purchase of the dining plan, the downside is that they expire at the end of the semester.
And so occasionally people end up with too many dining dollars at the end of the semester that they need to get rid of.
And so my plan, is to create a service that matches people who have excessive dining dollars, with people who are willing to pay real money for the food. Person with dining dollars buys the food, and the end customer pays the other student with real money, so it's a win-win.
One person converts their fake money into real money, and the second buys food at a discounted rate.
