** Two-sided confidence, exact method (N<=100)**

For sample sizes of 100 or less, the binomial exact method (Ref. 1) is used for calculating confidence intervals.
Given inputs of sample size (N), the number of defects found (Nd) and risk (α), which is one minus confidence, this tool iteratively seeks a solution by varying p_{U},
the upper confidence level for the proportion defective (p) in the left side of the equation, until the left and right sides are equal. The point estimate for the proportion defective is p=Nd/N.

The following equation is solved for p_{L} to obtain the to obtain the lower two-sided confidence level:

** One-sided confidence, exact method (N<=100)**

The calculation method for single sided limits are nearly identical to above, except all the risk is in either the upper or lower tail of the distribution. Therefore, α/2 is multiplied by two,
to represent all risk being in only one tail of the distribution.
The following equations are solved iteratively to determine the single-sided upper confidence limit (p_{U}) or single-sided lower confidence limit (p_{L}):

** Two-sided confidence, Wilson method (N>100)**

The following two equations (Ref. 1) are used to calculate the two sided confidence levels for cases where the sample size is greater than 100.
z is the variate value from the standard normal distribution such that, for two-sided confidence intervals, area in each tail of the standard normal distribution is α/2.

** One-sided confidence, Wilson method (N>100)**

To calculate single sided upper or lower confidence for cases where the sample size is greater than 100, the following two equations are used,
which are only different than the above equations in that z is not divided by 2, thereby representing all probability in only one tail of the distribution.

Toolkit Home

**Comments/Questions:**

reliabilityanalytics.com

- http://www.itl.nist.gov/div898/handbook/prc/section2/prc241.htm.
- MIL-HDBK-338, Electronic Reliability Design Handbook.
- http://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval
- http://en.wikipedia.org/wiki/Binomial_distribution
- O'Connor, Patrick, D. T., Practical Reliability Engineering.
- Bazovsky, Igor, Reliability Theory and Practice.
- Khan Academy, Margin of Error 1.
- Khan Academy, Margin of Error 2.
- Khan Academy, Confidence Interval Example.

All content and materials on this site are provided "as is" Reliability Analytics makes no warranty, express or implied, including the warranties of merchantability and fitness for a particular purpose; nor assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed; nor represents that its use would not infringe privately owned rights.