Top-down vs. bottom-up risk assessment: Consistent, contradictory or complimentary?

Identification of appropriate risk analysis techniques for railway turnout systems

“…The importance of ensuring how accurate and appropriate data are collected is vital. Given the subsystem levels of a railway turnout as the aim of the risk assessment study, it is expected to have two possible sources of data which might be used for the assessment: 1) data through the analysis of similar railway systems, such as crossing, and then allocation/contribution of failures to the subsystems of a turnout, and 2) data through elements and components of the subsystems of a turnout [47]. The latter is known as the bottom-up approach, while the former is the top-down approach.…”

“…Effort in deciding the structure of the study could be unrealistic. ASA`s recent study of a complex engineering system [47] revealed that a sound estimation might be achieved with the application of both to a study, and then the aggregate of the study outcomes and overall failure probability can be reached using techniques such as Monte Carlo.…”

Modules and transactions: Building blocks for a theory of software engineering

“…Software is one such domain, due to the existence of module hierarchy. Both terms, top-down and bottom-up, appeared in the software engineering literature in various contexts, such as, in relation to software development methods [31,28,2,33,37,9], process improvement [44], software reuse [30,29], product-line engineering [25], risk assessment [32], program comprehension [40], and software cost estimation [51,22]. The terms are also used in relation to software development governance [43]; however, the terminology did not gain an operational and commonly agreed upon exactness or demarcation.…”