To compute accuracy profile adopted by the SFSTP guide we should use Beta expectation as an interval, so I find another interval which is Beta content in the statistical field. I would like to know if it is possible to use it and what the difference between those intervals is.
Check this out:
Beta-Expectation and Beta-Content Tolerance Limits for Balanced One-Way ANOVA Random Model; Technometrics Vol. 26 No. 3;QICID: 9166 ,1984 pp. 251-254
http://asq.org/qic/display-item/index.html?item=9166
Assume you have a random variable X, and you want to know what values it will take. The answer is an interval, associated with a probability. For instance, you can say "X will be in [a, b] with probability beta". Then beta is the content of the interval.
If you knew everything about X law, the above sentence would be true. Then, the interval is called (at least in French) a "bet-interval" (not sure of the exact English term).
However, in general, you do not know exactly the X law, you just estimate it from a previous sample. In this case, the sentence is not true, you can just hope it is, because A and B are now random themselves. How to quantify how much you trust it?
If you say "In average, it will be true", then the interval is called a (beta-content) expectation tolerance interval. For the Gaussian case, it also correspond to the prediction interval for a single value of X. I'm not sure if this is true in general.
If you say "It will be true with probability gamma", then it is a gamma-confidence,, beta-content tolerance interval. It will be wider than the previous one.
Hope this is clear, and that English usage is not so different from the French one that I misused the terms (but please anyone correct if I did),
Best regards
In case you still have problems to understand I suggest to write to the following address:
E-mail: [email protected]
what i understand from this article and what i find that their is a difference between the two interval which is the way how to compute them but what i wish to know, when i should use Beta expectation and when i should use Beta content ?
thank for your explanation
best regard
Here on researchgate you have Robert Mee who has written:
Determination of sample sizes for setting [beta]-content tolerance limits controlling both tails of the normal distribution
Ask him
Best Béatrice
- Badges
- Science topic