Chapter navigation
Rigour in data, methods and judgement
Rigour turns this culture into practice by ensuring that decisions about data and methods are supported by assurance, evidence and documentation that can be tested and challenged.
Standard 6, Be Rigorous, reflects a straightforward expectation: official statistics should be built on suitable data and sound methods. Suitability depends on the question being answered, the context in which the statistics will be used, and the decisions they may inform. Quality therefore requires professional judgement about whether statistics are fit for purpose, rather than simple adherence to a method.
This judgement begins with understanding data. Producers should know where their data come from, what potential biases may exist and how changes in data collection might affect the resulting statistics. This can be difficult when data sources are complex, such as administrative data, linked datasets, new or innovative sources and methods. Understanding these sources of uncertainty helps producers decide what assurance is needed and how far the resulting statistics can support different uses. The task is not simply to access more data, but to use data responsibly and ensure the statistics remain relevant and reliable.
Rigour also means grounding methods in recognised professional standards and maintaining proportionate quality assurance throughout the statistical production process. Statistics should be validated and reviewed to ensure they continue to meet user needs, with comparisons to other evidence where possible. Quality is strengthened when producers remain curious about where weaknesses might arise, rather than assuming established systems will always perform as expected.
That includes testing the relationship between the statistic and the real-world phenomenon it is intended to represent. Producers should ask whether the estimate makes sense in the wider evidence landscape: what other data, intelligence or user insight supports it, and what might call it into question? Triangulating with other sources can help identify potential bias, changes in data behaviour or emerging quality concerns that would not be visible from the production process alone.
This kind of rigour depends on testing professional judgement as well as applying technical expertise. Producers need to make informed decisions, but they also need to test those decisions against evidence, alternative explanations and user challenge.
Viewed this way, rigour is more than a production process. It is a disciplined commitment to examining evidence carefully and making frank judgements about whether statistics are suitable for their intended uses.