WHERE IS THE FAULT IN FAULT-TREES

When branches of a fault tree are pruned, their probabilities are not fully transferred to the ''all other'' branch, We test 3 explanations for this underestimation of the ''all other'' probability: availabilit y, ambiguity, and credibility. In an experiment, we varied the credibi lity of a cover story and separately observed the generation of a faul t's causes to isolate availability, and the categorization of causes t o assess ambiguity. The results identify biased availability as a broa d threat to the validity of likelihood estimates. Ambiguity adds to th e problem whenever tree designers are unable to eliminate it from caus es or categories. Finally, though subjects had clear expectations for what constitutes a credible fault tree, none of the ''all other'' unde restimation could be traced to credibility factors. The discussion cov ers both underlying mechanisms and corrective techniques.