 |
Convergence to non-central t
curve Draws a Monte Carlo simulated curve converging
to a non-central t distribution curve. |
|
|
| n |
|
Sample size. |
| AQL |
% |
Acceptable quality level (%). |
| α |
% |
α, a rejection (Type I) risk. (See below.) • |
| ntr, jrepeat, klass |
|
No. of trials (samples), repeatability,
no. of histogram classes |
| tmin, tmax |
* |
Minimum and maximum abscissae for the graph. • |
| Show values |
|
Shows the coordinates of the graph. |
Draws a Monte Carlo simulated curve, which is a histogram,
converging to the theoretical density curve, a non-central t
distribution, as the number of trials (samples) tends to infinity.
In acceptance sampling, the “quality index” is
Q = t ⁄ √n, whence
k, the “acceptability constant”:
k = Qcrit. The t distribution applying
has parameters f = n − 1 and
δ = √n
Φinv(1−AQL).
Suggested risks, α, are given,
with the corresponding k (standard ANSI/ASQC Z1.9-1980).
* The default parameters
are suitable for the AQL’s presented. |
| n | α |
AQL (%): 1 |
1.5 |
| 5 | 10 |
k: 1.53 | 1.40 |
| 7 | 10 |
1.62 | 1.50 |
| 10 | 10 |
1.72 | 1.58 |
| 20 | 8 |
1.82 | 1.69 |
| 25 | 7 |
1.85 | 1.72 |
|
|
|
|
References:
• Resnikoff & Lieberman, 1957 (see elsewhere) |
