Competence and control are different
CREAM, the Cognitive Reliability and Error Analysis Method, relates human reliability to the context in which activity occurs. It distinguishes a person’s capability from the conditions that allow control of a particular situation to be maintained.
Analysis of human and organisational factors therefore does not end with confirmation of training. A competent person may receive incomplete information, incompatible objectives or insufficient time, affecting how the person observes, interprets and acts within that specific operating context.
Common performance conditions
The method examines conditions including organisation, working environment, interfaces, procedures, available time, simultaneous goals, experience and collaboration. These are considered for particular activities rather than assigning a generic rating to the whole company without distinguishing scenarios.
For example, an interface suitable for normal operation may become less useful when several related alarms appear. Analysis should explain what information the team needs and how its availability changes, rather than transferring an ordinary-operation rating without review.
Basic and extended analysis
CREAM includes a basic approach focused on context and an extended analysis addressing cognitive demands and functional failures. Merely adding checklist scores does not turn the result into a direct measurement of what will happen in the workplace.
Assessments require task knowledge, consistent criteria and an explanation of uncertainty. Where analysts interpret a condition differently, examine the evidence and scenario instead of concealing disagreement through an average that has no defensible methodological basis.
Preventive value and limitations
A useful application may identify information presentation, coordination or response time as the priority for improvement. That changes preventive decisions: not every problem can be solved by additional training or reminders for personnel to concentrate more carefully.
The method requires expertise and does not replace safe engineering design or overall risk analysis. Results should explain the context represented and when they cease to apply. They cannot diagnose individual psychological capability or justify comparing employees through a supposed personal reliability score. Changes to systems or operating demands require relevant assumptions to be revisited.
