...
Dan walked through a model at https://statswiki.unece.org/display/GSBPM/Clickable+GSBPM (requires login, but also may not be current), including a section that provided the overview of the whole process. The design / design frame & sample (2.4) section in particular is quite important for setting up a statistical programsurvey, whether the sample is probability based or not. We should reference this from IND with respect to concepts related to a statistical activity. Note that much of what we are concerned with in FIBO happen at the Analysis phase – because we are interested in the results of the processestimates, not so much the sausage making leading up to that point. individual records from acquired data. They are often too granular for public consumption. We are also interested in understanding the sources provenance of data used to generate the statistics. Some of this is straightforward for certain indicators, but for others it is more opaque. It really depends on the program – CPI is complicated, whereas some other statistics, such as the unemployment rate, are straightforward. Productivity is even more complicated than CPI, so having an understanding of the process may be useful for some business analysts who use those numbers.
GSBPM describes the approach from a high level, which may be important for some users of these statistics to know, so we should say something about this in the indicators ontology.
See https://statswiki.unece.org/display/GSBPM/II._The+Model#II._TheModel-_Toc375051196 and https://statswiki.unece.org/display/GSBPM/IV._Levels+1+and+2+of+the+Generic+Statistical+Business+Process+Model for the publicly available versions of what we walked through.