One important class of reliability data analyses
concerns the statistical identification of trends and
patterns inherent in the data over time. Following
Atwood et al., 1 a trend means a steady increase or
decrease over time in a reliability quantity of interest,
such as a failure rate (or probability). On the other
hand, a pattern means any deviation from a stable
state or condition resulting from some assignable
cause more fundamental than those producing the
mere randomness of the data while in the stable state.
An example might be the significant step increase in
the failure rate in a certain year due to the discovery
of a generic problem.