Issue #2-5: Paragraph 13

(Meeting 25th September, 2018)

• The addition of sentence might not be helpful here.
• In general, level of details in description of overarching process (OP) is quite imbalanced, this is something that Supporting Standards Group is intending to work on as an activity in 2019. Having elaboration on the two OPs only (in Section IV) might have misled readers. 
• Also need to look at other places in GSBPM doc. that could confuse or mislead readers, such as text in the green box in the diagram (this green box is for OPs, but it currently contains Quality Management/Metadata Management only); title of Section 6 (title is "Overarching Processes" but currently mainly talk about QM and MM only) as well as consider rearranging contents in paragraph 13 and paragraph 108.

Action: Alice and Manuel to draft proposal for descriptions of Data Management and Process Data Management respectively for Section IV. 

(Alice proposed following text for overarching process data management; 15th October, 2018)

Data management is essential as data are produced within many of the activities of the statistical business process and are the key outputs.  The main goal of data management is to ensure that data are appropriately used and useable throughout their lifecycle. Managing data throughout their lifecycle covers activities such as planning and evaluation of data management processes as well as establishment and implementation of processes related to collection, organisation, use, protection, preservation and disposal of the data.

How data are managed will be closely linked to the use of the data, which in turn is linked to the statistical business process in which the data are created. Both data and the processes in which they are created must be well defined in order to ensure proper data management.

Examples of data management activities include:

• Establishing a governance structure and assigning data stewardship responsibilities
• Identifying repositories in which to store the data
• Registering and inventorying data
• Classifying data according to content, retention or other required classification
• Determining retention periods of data
• Documenting the content and context of the data to ensure their use and interpretation as intended
• Securing data against unauthorized access and use
• Performing back-ups
• Safeguarding against technological change, physical media degradation, data corruption, and bit rot or decay
• Performing periodic checks to ensure that data integrity has been maintained (e.g., verifying checksums)
• Performing disposition activities once the data’s retention period has expired

(Feedback from ABS; 16th October, 2018) 

• It was noted (as an aside in the case of GSBPM) that "Data Management" could receive more coverage in GAMSO, similar to "Metadata Management" and "Quality Management"
• In the interest of alignment with Data Management beyond official statistics it was asked whether DMBOK (overview) could be cross referenced.
  
  Quality Management and Metadata Management could be considered covered elsewhere as overarching processes and some other segments are probably better covered in GAMSO,
• Other aspects covered by the ABS Data Management Framework that might be outside the current dot points in the draft overarching process description might be
• Broadly in the Data Modeling and Design segment
• designing data structures, and associated data sets, to make them as usable as possible for tools, business processes and consumers (potentially other statistical programs), and
• designing the flow of data through the statistical business process
• Broadly in the Reference & Master Data segment
• Managing a coherent view of providers/units and consumers of statistics (organisations and contacts), including interactions with them across Statistical Programs, updating Register information based on field feedback etc
• Maintaining other reference data used widely by systems such as
• Authoritative list of Statistical Programs and Cycles
• Authoritative exchange rate data
• Authoritative expression of geospatial information (including geometries/boundaries) that can be associated with statistical data

(Feedback from Danny; 16th October, 2018) 

Examples modified by Danny

• Governance, e.g., establishing a governance structure and assigning data stewardship responsibilities
• Architecture and modelling, e.g., design and implementation of IT architecture and processes
• Database and storage management, e.g. identifcation of repositories in which to store the data, database administration, business continuity planning
• Documentation, e.g. registering and inventorying data; classifying data according to content, retention or other required classification; determining retention periods of data; documenting the content and context of the data to ensure their use and interpretation as intended
• Legal aspects, e.g., securing data against unauthorized access and use
• Data ethics, e.g. evaluating moral problems related to data, algorithms (including artificial intelligence and robots) and corresponding practices (including responsible innovation and hacking), in order to formulate and support morally good solutions (e.g. right conducts or right values)
• Digital preservation, e.g., performing back-ups; safeguarding against technological change, physical media degradation, data corruption, and bit rot or decay
• Data integrity, e.g. performing periodic checks providing assurance about the accuracy and consistency of data over its entire lifecycle
• Disposition activities, e.g. once the data’s retention period has expired

(Meeting; 16th October, 2018)

• Data management: Alice will incorporate examples from ABS and Danny 
• Process data management: to be discussed when Manuel is available

(Manuel proposed following text for overarching process: process data management; 16th October, 2018)

Consists of the activities of registering, systematizing and using data about the implementation of the statistical business process (e.g. in a census or survey, data about how long the interview took to complete, whether the person used records to answer any questions, who responded to the interview). Process data–also known as paradata–can aid in detecting and understanding patterns in the data collected, as well as in evaluating the execution of the statistical business process as such. 

Sources:

https://www.census.gov/newsroom/blogs/research-matters/2017/04/paradata.html

http://researchaccess.com/2011/11/meet-the-data-triplets-data-metadata-and-paradata/

(Below version of Data Management description was included in GSBPM based on three considerations: 1. distinct from activities that could be covered by GAMSO; 2. keep the example list not too long; 3. start with action verb to be consistent with Metadata Management description; 21st November 2018)

Data management is essential as data are produced within many of the activities in the statistical business process and are the key outputs.  The main goal of data management is to ensure that data are appropriately used and usable throughout their lifecycle. Managing data throughout their lifecycle covers activities such as planning and evaluation of data management processes as well as establishing and implementing processes related to collection, organisation, use, protection, preservation and disposal of the data.

How data are managed will be closely linked to the use of the data, which in turn is linked to the statistical business process where the data are created. Both data and the processes in which they are created must be well defined in order to ensure proper data management.

Examples of data management activities include:

• Establishing a governance structure and assigning data stewardship responsibilities;
• Designing data structures and associated data sets, and the flow of data through the statistical business process;
• Identifying database (repositories) to store the data and administration of the database;
• Documenting the data (e.g. registering and inventorying data, classifying data according to content, retention or other required classification, etc.);
• Determining retention periods of data;
• Securing data against unauthorized access and use;
• Safeguarding data against technological change, physical media degradation, data corruption;
• Performing data integrity checks (e.g. periodic checks providing assurance about the accuracy and consistency of data over its entire lifecycle;
• Performing disposition activities once the retention period of the data is expired.