LogFAQs > #935141412

LinkPizza posted...

I said, from the beginning, that not everything is connected. For example, I said multiple times that the poll voters going down does not mean the user are necessarily going down. They CAN be connected, but it doesn't have to be. Which makes it a bad metric to count users. Which many people have also said.

But when front page metrics are down, and we see that the rest of the site is also down, that's when we say "The metrics show that overall site usage is down." That's how site activity measuring works.

LinkPizza posted...

Going board to board is better than guessing based on what you want it to be, though. And doing that also gets you a sample size. Tbh, it sounds like the reason you don't want to ask is because you know you have the possibility of being wrong. And while sample sizes aren't always right they are a way to gauge certain things in research. And when people do research, lots of them include sample sizes. You don't need to count every user. But counting some is better than puling a random number out of your ass because it matches what you want it to match. If you want people to actually listen to what you have to say or believe you, at least do more than make up numbers to fit your research.

A handful of people isn't good for anything. If I don't need them, I shouldn't go looking for them. But if I do need them, I need a significant amount of them. And going around to the boards wouldn't get me that. It would get me what I have here, the suggestion that the polls will be barren, but everything else on the site will be as active as it ever was. It's not likely. And if I could sample enough people, we would see what the likely truth is. But I don't need to sample enough people.