1. Mapping Data Production Processes to the GSBPM
Steven Vale
UNECE

2. Introduction
All processes that result in data outputs can be mapped to the GSBPM
Why do this?
Standardised documentation
Benchmarking
Quality assurance
First step towards common solutions for similar processes

3. Not a linear model Reminder
Sub-processes are not followed in a strict order
It is a matrix, through which there are many possible paths

5. Example 1 Australian project to improve production of prices data
Identified activities within GSBPM sub-processes
Some phases “out of scope”

7. Example 2 Another price index, this time for legal services in Norway
Note the difference in phases / sub-processes covered
E.g. no data integration or classification in this process

8. Price index for legal services
Quality Management / Metadata Management
Specify
Needs
Design
Build
Collect
Process
Analyse
Disseminate
1.1
Determine needs for information
2.1
Design outputs
3.1
Build data collection instrument
4.1
Select sample
5.1
Integrate data
6.1
Prepare draft outputs
7.1
Update output systems
General dissemination policy:
About the statistics
Tables in StatBank
Text & graphs
No integration carried out
Index & 4 sub-indices
Confidence intervals
Coefficient of variation
Tables & graphs
Eurostat regulation
Ministry of Justice
Trade association
Major companies
Standard first page
Test paper & electronic
Companies selected
according to sample plan
Load data into StatBank
5.2
Classify and code
1.2
Consult & confirm needs
2.2
Design variable descriptions
3.2
Build or enhance process components
4.2
Set up collection
No classifying or coding carried out
6.2
Validate outputs
7.2
Produce dissemination products
Check most important companies have replied. Iterative with 4.3
Check variation coefficient.
See 1.1
See Eurostat regulation
Agree on level of detail
Translation
Integrated system for editing and estimation
Sample register
Time plan
Preprinting, packing, sending, scanning etc.
5.3
Review, validate & edit
Daily statistics
About the statistics
Tables & graphs
Duplicates
Regional and product views
Discover outliers
Correct microdata
1.3
Establish output objectives
2.3
Design data collection methodology
3.3
Configure workflows
4.3
Run collection
6.3
Scrutinize & explain
7.3
Manage release of dissemination products
Packing & sending
Receive electronic data
Scan papir
Update received reg.
Send warnings & fines
Consult trade association.
Trade trends
Any previous periods
Other price indexes
Provided in regulation
Meetings with the various data capture divisions
5.4
Impute
Paper & electronic questionnaires
Scanning
Statistical calender
Approval routines
ssb.no
No imputation carried out
1.4
Identify concepts
2.4
Design frame & sample methodology
3.4
Test production systems
4.4
Finalise collection
6.4
Apply disclosure control
7.4
Promote dissemination products
5.5
Derive new variables & statistical units
Variables
Classifications
Establish population
Make sample plan
Stratification by size
Test micro-controls
Data from electronic &
paper questionnaires
Deliver data to processing
Daily statistics
Aggregation & publication level
Publication variables
Observation units = analysis units
1.5
Check data availability
2.5
Design statistical processing methodology
3.5
Test statistical business process
5.6
Calculate weights
6.5
Finalize outputs
7.5
Manage user support
Weights are established in phase 2
Weights are stored per survey
Final consistency checks
Supporting internal documentation
Choose companies (population) together with trade association
Micro controls
Weighting
Estimation
No field test or pilot
carried out
Web statistics
Mail or telephone
Contact 1.1 users
1.6
Prepare business case
2.6
Design production systems & workflow
3.6
Finalise production system
5.7
Calculate aggregates
Aggregates are calculated
Project plan
Finances
Scanning paper questionnaires
Electronic data capture
Documentation
Training
Meetings with 1st and 2nd line support
5.8
Finalise data files
Data files are finalised

9. Example 3
Irish tourism survey
Mapped GSBPM to 78 survey process steps

12. Example 4 Denmark – quarterly survey on employment in construction
Established, regular survey
Just phases 4 to 7

14. Example 5 Mapping business register processes to GSBPM
New Eurostat project:
Build up the capacity for using GSBPM and GSIM to document the national statistical business register processes
Describe national statistical business register processes by using GSBPM and GSIM
Grants will be available

15. GSBPM for business registers?
Business register maintenance is a continuous activity, not a single process
But BRs have:
Inputs “collected” from different sources
A sequence of processing and analysis
Outputs – statistics and sampling frames
Therefore BR maintenance can be seen as similar to other statistical production

17. Is this just an academic exercise?
No – there are practical benefits:
Standardisation of terminology
Standard framework for benchmarking
Facilitates use of common tools / methods
Efficiency savings
Tool for managing process quality

18. Conclusion
GSBPM can be applied to statistical business register maintenance
There is clear potential benefits in terms of shared knowledge, methods and tools

19. Exercise Map two processes to the GSBPM
Step 1 – determine which sub-processes are used, and in which order
Step 2 – For the “Process” phase, list the steps within each sub-process used