In general, the maintainability index to be demonstrated is the primary consideration in selecting a maintainability demonstration test method. Considerable savings in sample size can be obtained by use of sequential test procedures in preference to fixed sample sized tests. As a general rule, however, the sequential test should be used only when prior knowledge (e.g., from the prediction) indicates that the equipment may be much better (or worse) than the specified value.
The justification for use of the log-normal assumption for corrective maintenance time is based on extensive analysis of field data which have shown that the log-normal distribution provides a good fit to the data. However, in some cases it might be suspected that the log-normal assumption does not hold (e.g., equipment with a high degree of built-in diagnostics). When this occurs the log-normal assumption should be tested through the use of goodness of fit tests such as the Chi-square or Kolomogorov-Smirnov. If other distributions also do not fit the data, then a distribution-free method should be employed to ensure preservation of specified risks.