Hacker News new | past | comments | ask | show | jobs | submit login

> How does sampling theory support that your two experiences might be a trend? I got two flat tires in the last ten years. Does sampling theory say it's a trend? Statewide? Worldwide?

I don't run with startup/entrepreneurial types, so for this to come up in the way it has with my limited exposure to these folks, I consider that at least somewhat significant. I am not saying it is definitely a trend. But I do remember from my social science labs that ballpark-accurate statistical inferences can be made from surprisingly small sample sizes if the population sampled is broad enough. Do I just happen to find myself around locationally-anxious ex-Californian business owners? Who knows.

> Absolutely, but let's remember that much of the narrative has been pushed by a political faction for years

Ah yes, politics' icy tendrils (tentacles?) seem to molest everything they come across, and even some things they don't. All I can say is that I am not trying to push a political narrative, and that the people who are should kindly jump off a cliff.




I think those small samples need to be larger than that, and they have to be representative samples.

> I am not trying to push a political narrative, and that the people who are should kindly jump off a cliff.

I am not at all saying that you are. I'm saying that the narrative affects the situation. I agree about the cliff!


9 people gets you a significant portion of the confidence that a larger survey provides.


? It depends on the size of the population. I don't believe surveying 9 people will tell you anything about public opinion among the ~325 million Americans.

And again, it must be representative. Should I ask 9 friends of mine?


No, unless you are trying to draw conclusions limited to your friends. Similarly, two flats in 10 years might be significant for you and your driving, but not tires as a whole.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: