4 ms·
>> How much data is enough? > How much data do we have? Say we have X. Do we need 2X then? Because that's a request that never ends.
by euyyn 10y ago
>> How much data is enough?
> How much data do we have?
Say we have X. Do we need 2X then? Because that's a request that never ends.
- marcosdumay 10y agoI'm saying that, as far as I know, X = 0. 2X won't be enough either, and you better work on getting some reasonable evidence that your data is impartial too.
- eanzenberg 10y agoUntil it's statistically significant. And if you then ask "well do we wait for 2 sigma or 5 sigma" then you're just playing devil's advocate. Peer-reviewed science isn't perfect but it moves towards better understanding of nature.
- Kalium 10y agoPeer-reviewed science isn't perfect. The point I was implying previously is that to many people, this means it can never provide a satisfactory answer, because the unspoken requirement is perfection.
- eanzenberg 10y agoRight but even for life/death science such as in the medical community, drug development, space science, etc., those people who oppose any progress of science will never be satisfied with the current-day process.
- Kalium 10y agoPrecisely. I've been implying that it's a waste of effort to try to please those people. They cannot be pleased, and there are other uses for those resources.