i think a really funny project that a statistics professor could have their class do is like. put a bunch of random, patently untrue demographic statements into a hat. “the most popular tv show among white men ages 24-27 is Bluey.” “the majority of business majors are middle children.” “bisexual women love hot chips.” and each student picks one out of the hat and you gotta like. design a whole study and survey a group of people to specifically achieve that result. you have to prove it true. by whatever means necessary. you have to construct the most biased study possible and wrangle in your exact demographic to make that statement a statistical reality. i think people would learn a lot.
CLARIFICATION: students would be forbidden from recruiting participants with any communication that contains more than 1 demographic qualifier!!
e.g. they can put out an ad for “female study participants” but not for “bisexual women who like hot chips”
this forces them to design a screening process that narrows it down to their desired group. “yeah sorry the survey takes 2 hours, and it’s right during lunch, but we do have an unlimited supply of hot chips you can eat.” and then occasionally a beautiful butch woman walks by the windows idk im still brainstorming