2 ms·
Let's say I measure a rope many times, and come up with a mean value and a 95% confidence interval for the length of the rope. That is, assuming my "experiment"
by scott_s 17y ago
Let's say I measure a rope many times, and come up with a mean value and a 95% confidence interval for the length of the rope. That is, assuming my "experiment" is constructed correctly, I say that the true length of the rope lies within that range, with a 95% confidence. That's an objective confidence.
But in order to come up with that number, I had to assume my experiment was correct. I very well could have had a systematic error in the experiment such as misusing my ruler, accidentally holding the rope such as to artificially shorten it, or completely misunderstood the concept of length. My confidence that my experiment is correct is both independent of the confidence interval I reported, and not quantifiable.
- dejb 17y agoWhat about measuring the speed of light? or measuring the distance from the earth to sun? Where do these fall in your neat divisions between subject and objective? Is there a certain class of proposition that is too complicated to be called objectively true? Are all medical theories, for example, simply subjective and have no real meaning in terms of predicting likely future outcomes? Any statement or theory that does not have predictive power is meaningless. Predictive implies probability. Probability implies odds. Put up or shut up up.