One Size Doesn’t Fit All: Administrative Data Quality Frameworks for Production of Official Statistics

Abstract

Background with rationale
The use of administrative data is key to achieving the UK Statistics Authority’s strategy of Better Statistics, Better Decisions. Integrating administrative data into official statistics can benefit policy decisions by allowing the possibility of greater granularity and improved timeliness in outputs, while delivering efficiency gains and reducing respondent burden. Quality assessment and communicating uncertainty of administrative data sources is critical to their effective integration into official statistical outputs.

Main Aim
This presentation will discuss the main challenges of quality assuring statistical outputs containing administrative data. The differences in existing quality frameworks and identified quality metrics will be discussed. In addition, the presentation will cover the need to tailor quality assessment to answer a specific research question that an identified source is being used for and the considerations required.

Methods/Approach
A comprehensive literature review was carried out, bringing together existing quality frameworks and metrics from National Statistical Institutes (NSIs) and academia for production of statistics using administrative data sources.

Results
The main challenges and considerations faced when quality assuring outputs produced using administrative sources have been identified. The quality requirements for different outputs across social, business and census statistics were summarised and a general quality framework for admin data developed. This framework draws on international best practices for use in the UK statistical system.

Conclusion
Integrating administrative data presents challenges can’t be solved by a one-size fits all framework. Through unifying available guidance, an adaptable quality assurance methodology has been created, enabling the use of public data for the public good.

Background with rationale

The use of administrative data is key to achieving the UK Statistics Authority’s strategy of Better Statistics, Better Decisions. Integrating administrative data into official statistics can benefit policy decisions by allowing the possibility of greater granularity and improved timeliness in outputs, while delivering efficiency gains and reducing respondent burden. Quality assessment and communicating uncertainty of administrative data sources is critical to their effective integration into official statistical outputs.

Main aim

This presentation will discuss the main challenges of quality assuring statistical outputs containing administrative data. The differences in existing quality frameworks and identified quality metrics will be discussed. In addition, the presentation will cover the need to tailor quality assessment to answer a specific research question that an identified source is being used for and the considerations required.

Methods/Approach

A comprehensive literature review was carried out, bringing together existing quality frameworks and metrics from National Statistical Institutes (NSIs) and academia for production of statistics using administrative data sources.

Results

The main challenges and considerations faced when quality assuring outputs produced using administrative sources have been identified. The quality requirements for different outputs across social, business and census statistics were summarised and a general quality framework for admin data developed. This framework draws on international best practices for use in the UK statistical system.

Conclusion

Integrating administrative data presents challenges can’t be solved by a one-size fits all framework. Through unifying available guidance, an adaptable quality assurance methodology has been created, enabling the use of public data for the public good.