Table of percentiles for Regression with Life Data

The table of percentiles provides the age by which a percentage of the population is expected to fail. Use these values to determine whether your product meets reliability requirements, or to compare the reliability of two or more product designs.

The table of percentiles consists of the following columns:

Percent, which is the percentage of a population that is expected to fail.
Predictors, which indicate the combination of factor settings for that prediction.
Percentile, which is the age by which a percentage of the population is expected to fail.
Standard error, which is a measure of the variability in the estimate.
95% confidence interval, which provides a range of values that are likely to contain the true value of the parameter.

Use these values only with an appropriate model for the data. If the model fits the data poorly, then these estimates may not be accurate.

Example output

Table of Percentiles

				Standard Error	95.0% Normal CI
Percent	Design	Weight	Percentile	Standard Error	Lower	Upper
1	Standard	5.0	134.911	17.6574	104.385	174.363
1	Standard	7.5	117.113	16.0279	89.5591	153.144
1	Standard	10.0	101.663	16.3830	74.1295	139.423
1	New	5.0	273.214	36.8022	209.819	355.763
1	New	7.5	237.171	32.6878	181.028	310.726
1	New	10.0	205.882	32.8675	150.568	281.518
5	Standard	5.0	178.749	16.9676	148.404	215.300
5	Standard	7.5	155.168	14.1107	129.836	185.443
5	Standard	10.0	134.698	15.4568	107.568	168.670
5	New	5.0	361.994	36.0778	297.761	440.084
5	New	7.5	314.239	28.8741	262.450	376.247
5	New	10.0	272.783	30.6102	218.928	339.887

Interpretation

For the compressor data, the table of percentiles includes the following results:

Row 1 and Row 7 refer to the combination of a standard design and a projectile weight of five pounds. For this combination, 1% of failures are expected to occur before 134.911 hours, and 5% of failures are expected to occur before 178.749 hours.
Row 4 and Row 10 refer to the combination of a new design and a projectile weight of five pounds. For this combination, 1% of failures are expected to occur before 273.214 hours and 5% of failures are expected to occur before 361.994 hours.

These results indicate that a compressor case with the new compressor design survives longer than a compressor case with the standard compressor design.