(Feedback from ONS, 2 October, 2017)

Describe issue or need

Analyse Phase is too output-oriented. Maybe call it ‘primary analysis’ or ‘estimation’….

1) This makes it not generic enough (it ignores analysis that could relate to insight on earlier phases)

2) Assumes "data files" have been finalised

3) Forces GSBPM to be linear (i.e., Analyse can only happen after Process)

Phases impacted

6 - Analyse

Suggested Improvement

Suggest the scope of Analyse Phase should be increased to include initial exploration of a new or existing data source (for example, activities related to "playing with" or using a "sandpit" in order to understand a big data source) as well as ongoing analysis of data during processing or derivation of analytical insights during operations.

In practice, this creates two Analyse streams:

  • A process that applies an agreed methodology to a data source
  • A more flexible activity around generating insight about a data source.

This change would make it easier to mix processing and analysis activities as well: sequences such as (integrate data, analyse integrated data, process, analyse results, process, analyse and explain, publish) become feasible.

Analysis in this case also becomes an activity that can support the Design phase by allowing prototyping and generation of designs.

  • No labels

1 Comment

  1. InKyung Choi

    Revision Group Meeting 24 April, 2018 

    • GSBPM is non-linear, the position of phases in diagram is not the order of process, earlier phases can be iterated after the analysis
    • There can be more text for clarification, to broaden the idea that the phase is not only for the final output but also can be used as input for other phases or sub-processes and can be iterative (e.g. analysis from new or/and experimental data sources can be fed back to design/process/etc.) and there can be "preliminary" analysis in earlier stages (e.g. Design or Build) with new data sources
    • However, GSBPM is for regular business process, tasks already made mainstreamed, it might be risky to try to cover all activities like experimental data
    • The sentence "made ready for dissemination" in paragraph 84 seems imposing linearity, this should be changed. 
    • We will come back on the name of the phase (Analyse) later when we look at the overall issues.