"Researchers used a verbal joke test developed in 1983 and used in other humor studies. Mak added a new element, though, by showing participants cartoons from the Ferd'nand comic strip, and asking them to choose between four panels to locate the funny ending. Three of the choices for each cartoon were the wrong ones, created by an artist for the study.
"This wasn't a study about what people find funny. It was a study about whether they get what's supposed to be funny," Carpenter said.
I think it's a serious flaw to use the Ferd'nand strip. I get that they were asking people about what's supposed to be funny rather than what is funny, but seriously, sometimes it's hard to tell with that comic! And I've always thought that. Even before I got old. ;)
Also, obviously, I didn't see any of the researchers' statistical analysis, but young people doing 6% better on verbal jokes and 14% better on comics doesn't seem very significant when you're talking about a study population of only 80 people...
As for the study size, I seem to recall from Stats 101 that, if you have a homogenous population, you can extrapolate results from a sample as small as 30 with high reliability. I don't know the features of this sample was, but 80 is not necessarily too small.
Now, see, THAT was funny!
Here's the tricky bit, though. The article says,
I think it's a serious flaw to use the Ferd'nand strip. I get that they were asking people about what's supposed to be funny rather than what is funny, but seriously, sometimes it's hard to tell with that comic! And I've always thought that. Even before I got old. ;)
Also, obviously, I didn't see any of the researchers' statistical analysis, but young people doing 6% better on verbal jokes and 14% better on comics doesn't seem very significant when you're talking about a study population of only 80 people...
At least they didn't use Ziggy.
As for the study size, I seem to recall from Stats 101 that, if you have a homogenous population, you can extrapolate results from a sample as small as 30 with high reliability. I don't know the features of this sample was, but 80 is not necessarily too small.
ha ha i guess
Post new comment