Quality vs quantity: Stovepipe better than DWH? COE on S-DWH Wiesbaden 23.11.2016
The European Statistics Code of Practice Five principles covering statistical processes Sound methodology Appropriate statistical procedures Non-excessive burden of respondents Cost effectiveness 10 May 2019
The European Statistics Code of Practice Five principles covering statistical processes Sound methodology Appropriate statistical procedures Non-excessive burden of respondents Cost effectiveness Five principles covering statistical outputs Relevance Accuracy and Reliability Timeliness and Punctuality Coherence and Comparability Accessibility and Clarity 10 May 2019
Quality in ‘Best Practice’ cases Best practice cases emphasize the statistical processes Example: Statistics Finland’s planned costs and savings for S-DWH project 10 May 2019
Quality in S-DWH manual S-DWH manual deals mainly with quality of statistical processes Metadata chapter: 1.2.5 Metadata of the S-DWH layers Metadata chapters: 1.3.3 – 1.3.6 Metadata functionalities by layers Governance Chapter: 2.1 Governance of the Metadata Implementation: 2.1.2 Quality requirements for the statistical data There is not much about quality of statistical outputs 10 May 2019
Quality in S-DWH manual S-DWH manual deals mainly with quality of statistical processes Metadata chapter: 1.2.5 Metadata of the S-DWH layers Metadata chapters: 1.3.3 – 1.3.6 Metadata functionalities by layers Governance Chapter: 2.1 Governance of the Metadata Implementation: 2.1.2 Quality requirements for the statistical data There is not much about quality of statistical outputs ... until today! 10 May 2019
Group Work We’d like you to form into 5 quality groups, and discuss (a) how your quality dimension(s) can be measured in a DWH (b) how DWH & stovepipes perform re your quality dimension(s) Once you’ve had your discussion (~ 20 minutes) we would like each group to present feedback to everyone (~ 5 minutes per group) GO! 10 May 2019
Relevance Who are current/potential users What are their needs Does output meet needs 10 May 2019
Accuracy & Reliability Sampling error Non-sampling error 10 May 2019
Timeliness & Punctuality Production time Frequency/punctuality of release 10 May 2019
Accessibility & Clarity Needs of analysts Assistance to locate data Dissemination format 10 May 2019
Coherence & Comparability other statistics/data in same domain or at different frequencies Comparability over time space domains 10 May 2019