Sensitivity Testing and Analysis Course

Instructor: Dr. Barry T. Neyer

The Sensitivity Testing and Analysis Course is an explanation of the test and analysis methods used in conducting sensitivity tests (such as the Neyer test, Langlie test, and Bruceton test). The class covers:

• What can be learned from a sensitivity test.
• How to perform a test.
• How to analyze and interpret the results.

The course covers both the basic theory and practical aspects of testing and analysis. Upon completion of this course, the students will be able to use software to conduct and analyze a wide variety of sensitivity tests.

This is a comprehensive course; for all technicians, professionals, and managers in development, production, quality, or test fire organizations who design, build, or test components that are destructively tested.

This course can be extended into a detailed course suitable for professionals involved in testing. The extended course is devoted to comparison of different test and analysis techniques through simulation. Students will also receive one on one instruction for testing and analysis of components of interest to students. All components to be tested must be supplied by the students.

Sensitivity Course Outline:

1. What is a sensitivity test?
2. Review of Statistics for Normal Distribution
   • Resistor example
   • Normal (Gaussian) probability curve
   • Mean. Standard deviation of population
   • Other parameters of the population
   • Likelihood functions
   • MLE
   • Variances and covariance of estimates
   • Confidence Intervals
   • Cramer-Rao Theorem, Asymptotic Confidence Intervals
3. Sensitivity test statistics
   • What is measured?
   • Statistics the same, measurements and computations different
   • Use of Likelihood function
   • Confidence Levels
   • Cramer-Rao Theorem, Asymptotic Confidence Intervals
   • Efficient test design
4. Efficient sensitivity testing
   • Egg drop test motivation
   • Explanation of new test
   • Can you do any better?
   • How to determine the sample size
   • Hands-on running of the software
5. How to use the software for testing
   • Mechanics of SenTest^TM
   • How to determine limits of Mu and Sigma
   • How to determine appropriate test hardware
   • Running SenTest^TM with simulated components
6. How to use the software for analysis
   • History Plot (Xs and 0s)
   • Contour Plot
   • Numerical Analysis
   • Probability Plot
   • Multi Contour Plot
   • Different Populations
7. Special Topics
   • What to do if no overlap
   • How reliable are no-fire and all-fire levels?
   • Adaptability of the Optimal techniques
8. Test
   • Comprehensive test of sensitivity test and analysis statistics

1. Second Half
2. Comparison of the various test techniques
   • Neyer D-optimal test
   • Neyer c-Optimal test
   • Neyer G-Optimal test
   • Neyer E-Optimal test
   • Original Neyer test design
   • Bruceton test
   • Robbins-Monro test
   • Langlie test
   • Adaptive Langlie test
   • Probit test
3. Comparison of Likelihood Ratio and other analysis methods
   • Likelihood Ratio Test
   • Asymptotic Confidence Intervals
   • Langlie Analysis
   • Bruceton Analysis
   • Function
   • Simulation
4. Test Student Supplied components (Optional)
   • Determination of initial parameters for components
   • Selection of tester
   • Conduct the test.
5. Analyze Student Supplied components (Optional)
   • Plot results
   • Compute confidence intervals
   • Compute No Fire and All Fire levels
   • Plot probability levels
   • Compare the results of two samples

Contact Neyer Software at (513) 777-6672, info@neyersoftware.com, 7275 Willowood Dr., Cincinnati OH 45241-3703

Contact us at Neyer Software

Return to Neyer Software Home Page