No CrossRef data available.
With Examples on Measurement of Aero-Engine Reliability
Published online by Cambridge University Press: 28 July 2016
The subject of statistics is so vast that many thousands of volumes have been written on it and it would be lunacy to try to give any complete outline in a brief paper. Even by confining the field to that range of data arising from aeronautical engineering the limit seems to recede into specialist abstractions rather than narrow as generality is jettisoned. At the outset it may be well to consider these apparent platitudes because many (most?) otherwise well-informed people are dogmatic in their belief that statistics are merely applied arithmetic. No doubt the same kind of thing was said before any distinction was drawn between a tally-clerk and a qualified accountant. The assertion contains too much fundamental truth for easy rebuttal: nevertheless there does exist both a science and an art in statistics, and arithmetic is only one important contribution to the exercise of either.
* If probability of failure = 10% = p and probability of non-failure=90% = q and the size of samples selected = 10 = n, then expansion of the binomial (p + q) n gives a probability distribution showing the chance (as a percentage of all possibilities) of various numbers of defective engines turning up among samples of ten selected at random.
Defective engine among the 10 | Chance % of such an answer |
---|---|
0 | 34.9 |
1 | 38.7 |
2 | 19.4 |
3 | 5.7 |
4 | 1.1 |
5 | Less than 0.2 |
6 | Less than 0.1 |
7 | Less than 0.001 |
8 | Less than 0.0001 |
9 | Less than 0.000,001 |
10 | Less than 0.000,000,1 |