START is a Reliability Information Analysis Center current awareness periodical. Each START document provides a 'jump-start' on a topic of immediate interest to individuals in the reliability, maintainability, quality, supportability, or interoperability fields. In addition to a concise write-up, each installment includes comprehensive lists of contacts, bibliographic references, and other sources for further information and action.

RelTIQUES are a Reliability Information Analysis Center technology publication that goes into more detail on specific topics than our regular START sheets. Each RelTIQUE will include more detailed technical topics of interest to reliability, maintainability, quality, supportability, or interoperability students and practitioners.

RIAC START Sheets by Category

     Reliability

     Maintainability

     Quality

     Supportability

     Interoperability

RIAC RelTIQUES

     RelTIQUES

Regression Analysis

Regression modeling is probably the most widely used (and misused) statistical tool in the analysis of data. Assume that we have two or more variables, that they are associated (correlated), and that the variable of interest is difficult to assess. However, the other variables are cheaper, easier or faster to obtain. Then, it is appropriate to use regression to find a model that expresses the variable of interest, as a function of all the others.

---

Understanding Exponential Sequential Tests

The actual sample size “n” required in testing and confidence interval (CI) derivation is of tremendous importance for practitioners. For example, sample size carries with it a price tag in time, resources or both. In industrial applications, neither of these is plentiful.

When samples are taken all at one time, it is called single sampling. The problem of calculating the sample size for deriving a general CI has been discussed in Reference 2. Samples for acceptance testing were presented in Reference 3. Censored samples have been discussed in Reference 4. All of these cases, however, only treat the situation where fixed samples of pre-determined sizes are taken, all at one time.

An alternative consists in taking the samples in multiple stages and assessing their results at each stage. This allows the possibility of stopping the process and reaching an early decision, if certain conditions are met. For example, if the start data are clear-cut in favor of (or against) the hypothesis, then curtailing the test can save significant time and resources. Such is the case where samples are taken in successive stages, according to the assessment of results obtained from the previous sampling stages. This is known as “multiple sampling”.

---

The RIAC START Sheets are available in PDF format.
If you do not have Adobe Acrobat Reader please download it here.