AudioBanter.com - View Single Post

(Thomas A) wrote in message ...

Yes, but still, if the difference is real and miniscule it could
introduce incorrect answers even if there is specific training
beforehand. If there would be an all or nothing thing, then the result
would always be 100% correct (difference) or 50% (no difference).
What if the answers are 60% correct?

You would still need far fewer than 500 trials to get a statistically
significant result. For 165 trials, 99 correct, which is 60%, would be
statistically significant at the 99% confidence level. If you were
willing to settle for a 95% confidence level, you would need even
fewer trials.

And remember, even if you don't get a statistically significant
result, you still can't conclude that the difference is inaudible. So
you wouldn't get an incorrect result; you'd get an inconclusive one.

bob