Q&R Statistics - Nonparametric Test Sample Size

[bradley]

Nonparam Sample Size

Calculates the sample size needed for a nonparametric reliability test to demonstrate a given reliability at a given confidence level for a given time. The reliability is only demonstrated if there are no failures during the test.


Variables

Confidence = 95.0 %
Suffix = %
Display to 1 decimal places

The confidence to be obtained. The confidence is 1 - (the probability that the given reliability might be obtained due to luck). Example: 95.0 %

Reliability = 0.95
Display to 3 decimal places

The reliability to be demonstrated for a time equal to the test length. Example: 0.95


Results

# units needed = 59
Display to 0 decimal places
Formula = ceiling(ln(1-(Confidence/100))/ln(Reliability))

The number of units to put on test. Example: 59


Notes

Use to determine the sample size needed for a zero-failure nonparametric reliability test.

The test is only successful at demonstrating R reliability at a time equal to the test length at C% confidence if none of the units fail during the test.

All units are assumed to be tested for the same length of time. The test time is set in advance
(i.e., do not test until a failure occurs and then set that as the test time).