There's been a lot of slamming going on here about this page, what its good for, etc.
First off, the url ends in oddtodd just because the idea actually started on the oddtodd forums. Beyond that, it's completely seperate. Yes, I realize the report pages offer little information statistically - at least at this point - I know that comparing one's unemployment and salary
There's been some criticism that I don't say that the tables don't mean jack (yet). A few things in my defense - I sent the link to a few yahoo groups. In each of those cases, I pointed that out. It also says so in the help, and up until recently it said so on the main page. I also figured that people have a brain and can figure out that if you are comparing yourself to 7 people....
Anyway, the site's mostly for fun with some statistical stuff being pulled out. IT's a general audience site. Someone here suggested I just list correlation coefficients. Common, would anyone understand it?
Finally, it's much harder to calculate correlation coefficients on the fly - simple everyday "I can understand what that means" figures like averages and maximums are much easier and people actually know what they are.
Does that mean that I won't try to use the questions for some "real" stats? No, I do plan to, but I'm not able to run SPSS regressions on the fly - mostly cause I wouldn't know where to start to make it work.
Anyway, thats it. Comments / thoughts welcome.
Slashdot Mirror
There's been a lot of slamming going on here about this page, what its good for, etc. First off, the url ends in oddtodd just because the idea actually started on the oddtodd forums. Beyond that, it's completely seperate. Yes, I realize the report pages offer little information statistically - at least at this point - I know that comparing one's unemployment and salary There's been some criticism that I don't say that the tables don't mean jack (yet). A few things in my defense - I sent the link to a few yahoo groups. In each of those cases, I pointed that out. It also says so in the help, and up until recently it said so on the main page. I also figured that people have a brain and can figure out that if you are comparing yourself to 7 people.... Anyway, the site's mostly for fun with some statistical stuff being pulled out. IT's a general audience site. Someone here suggested I just list correlation coefficients. Common, would anyone understand it? Finally, it's much harder to calculate correlation coefficients on the fly - simple everyday "I can understand what that means" figures like averages and maximums are much easier and people actually know what they are. Does that mean that I won't try to use the questions for some "real" stats? No, I do plan to, but I'm not able to run SPSS regressions on the fly - mostly cause I wouldn't know where to start to make it work. Anyway, thats it. Comments / thoughts welcome.