The analysis and comparison of start-up demonstration tests

Two new start-up demonstration tests are introduced as alternatives to a previously studied one. The first test is based on total successes and total failures (TSTF) while the second is based on consecutive successes and consecutive failures (CSCF). The probabilistic analysis of these tests in the independent and identically distributed case uses a Markov chain approach to avoid the complexities of the probability generating function approach. We discuss advantages of these new tests in certain situations. Practical guidance on choosing tests, estimation, and comparisons of the various criteria are studied. The Markov chain approach can be easily extended and generalized to study the probabilistic analysis of the non-i.i.d. case. (c) 2007 Elsevier B.V. All rights reserved.