Meeting notes - July 9th, 2024

[Chris-ECE]

Under the feedback for phase 2, it was decided that:

Subprocess 2.2 (comment from Sweden)

• The suggestion about mentioning contextual metadata in 2.2 is already addressed as the subprocess description refers to metadata, which would include also contextual metadata.

Subprocess 2.2 (comment from Hungary)

• As discussed in the previous call (in terms of differences between Phases 1 and 2 in relation to variables), it is the unit type (or statistical unit) that changes a concept into a variable, so we accordingly changed the name of 2.2 to “Design Variables” rather than “Design variable descriptions”.
• The proposal from Hungary suggests in 2.2 designing metadata for the entire statistical business process, but UNECE suggests keeping the scope of 2.2 to remain focused on variables (as there are lots of different things one might wish to collect metadata for, like workflows, which are the domain of 2.6). The introduction to the Design phase already says that the Design phase specified all relevant metadata.
• UNECE suggests avoiding saying that there is a particular subprocess of GSBPM where all metadata is designed, especially as “Managing Metadata” is an Overarching Process, and metadata is also mentioned in other subprocesses (e.g. 2.1).
• Cory agreed, and mentioned that we don’t just design metadata, but may reuse metadata designed elsewhere. He said it would be good to include mention of standards and corporate policies for metadata within the overarching processes. InKyung mentioned that one decision that has been made is to put the Overarching Processes towards the start of the GSBPM document, and she said that a mention could be made there about following corporate policies and standards. On this basis, some text was inserted near the start of the Metadata Management text in the Overarching Processes.

Subprocess 2.2 (comment from Mexico)

• For the comment about inserting text to clarify that the aim of following standards is to facilitate integration between the business processes, UNECE suggested that there are many other benefits of following standards (of which integration is just one) such as aiding communication or sharing the development of software tools, and that the description of subprocess 2.2 might not be the right place to mention this (as there might be calls to expand the text to mention all benefits of using standards).
• UNECE initially proposed to formulate a more general text to insert somewhere in the introduction of the document, however, upon further reflection would propose that an explanation of the benefits of adopting standards lies outside the scope of the GSBPM document. (Action: To discuss.)

Subprocess 2.2 (comment from France)

• It was clarified that this comment was about defining relationships between variables (for example if multiple survey questions are needed to calculate a particular variable, or to trace how different instances of a particular variable are calculated through the different process stages, which might involve imputation, etc.)
• Florian agreed that defining the linking/relationships between variables might be spread among various parts of the design phase, so might not be worth mentioning it within a particular subprocess description. Florian suggested putting it in an overarching process. Juan suggested including something within Data Management. A review of the existing text for the Data Management Overarching process was agreed to be sufficient to cover this aspect.

Subprocess 2.3 (comment from Hungary)

• This comment related to multi-mode design of surveys. UNECE made a proposal for some extra text for inclusion in the description of subprocess 2.3. “The collection may employ a multi-mode methods which requires additional considerations such as design of sequencing or different instruments for different modes.” This was agreed during the call.

Subprocess 2.3 (comment from Ecuador)

• This comment suggests including “instruments design to backup and protect data and design of the process of extraction, transformation and loading (ETL) for the system of database management”.
• It was noted that design of IT systems would pertain to subprocess 2.6 (Design production system and workflow).
• Juan suggested that the specific issue of ETL is too detailed to mention within the description of subprocess 2.3, but that more broadly, this comment might relate to using data from different sources (including administrative or unstructured data). Since one still designs acquisition of data such as administrative data, it was suggested to make this clearer in the subprocess description. While the text already makes a number of references to e.g. admin data, it was decided to make this aspect clearer by removing (possibly distracting) references to census and survey in the first sentence of that subprocess description.

Other matters

• It was agreed that one consideration of the phase-level comments was concluded, the tracked changes in the working document would be cleaned up so that a clean copy of the document can be sent to others for review.