In this paper, operators' cognitive appraisal of sensor-based alarm systems is examined. To date, researchers have followed an inductive approach to alarm reactions, proposing trust-based theories to account for variability in alarm reactions. Deductive reasoning has likewise led researchers to collect data in support of various theories. Existing theoretical frameworks assume a common precursor to operator belief systems: firsthand experience with alarm systems. However, humans are frequently exposed to signals with which they have limited knowledge. The construct of alarm "faith" is proposed to account for initial differences in alarm responding prior to experience. Implications of the constructual distinction between faith and trust are discussed with consideration given to task training applications.
展开▼