Quote:
Originally Posted by Spacemonkey
Quote:
Originally Posted by peacegirl
Quote:
Originally Posted by Angakuk
Quote:
Originally Posted by peacegirl
Until there is more testing done (that is reliable and replicable), I will continue to believe he was right. That sounds very reasonable and rational to me.
|
How will you know when enough reliable and replicable testing has been done? |
When the variables of the test are controlled such that there can be no statistically significant errors, or bias. |
And how will you be able to tell the difference between there really being no statistically significant errors or bias affecting the results, and there being no statistically significant errors or bias affecting the results that you can identify or detect (which is where you are now)? |
Why are you turning this into something I contrived Spacemonkey? A test with good controls is statistically significant but the tests I have seen have not been reliable [in my opinion]. A statistically significant test would not have any underlying bias which could easily interfere with the validity of the results in order to prove the original hypothesis [or theory] true. I'm not saying this is done purposely, but it's very tempting to misread the results of a test you want to be true. We can't avoid all bias, so we have to be vigilant in our effort to remove as much bias as we can. This means setting up the experiment in such a way that it doesn't lean toward the results we are hoping for.