Statistical processes can be complex. There are plenty of opportunities for systemic or unexpected errors to arise throughout the production process: in collection design, data capture, data processing and transformation, and calculation of statistics. Typically, statistical products are subject to formal clearance processes before being publicly released and disseminated. However, if problems are found at clearance time - or worse still, after release - there is little time and great pressure to find the source of the errors and make corrections.
Quality Gates are an initiative to help manage the quality of statistical processes. They assist in the identification of identify any problems close to their origin in order for them to be fixed in a timely manner. In this talk, Paul will explain the key components of effective quality gates, work through an example of a gate, and highlight their benefits.
This seminar should be of particular interest to staff involved in the production of statistics.
Paul Schubert - Assistant Statistician, Statistical Services Branch, Australian Bureau of Statistics.
Video Presentation (streaming) | Power Point Presentation (697KB)
Return to National Statistical Seminar Series