Quality Assurance Principles

After reading this Circular, you will understand the quality assurance principles that apply across all ocean accounting work. These principles cover data quality dimensions, uncertainty categories, and documentation standards, and are referenced by all other circulars in the Technical Guidance.

Essential prerequisites:

• TG-0.1 General Introduction to Ocean Accounts — for the conceptual framework and the three-domain structure of Ocean Accounts

Helpful background:

• TG-0.2 Overview of Relevant Statistical Standards — for the SEEA EA quality dimensions and statistical standards referenced throughout

3.1 Purpose of Quality Assurance

Ocean accounts integrate data from diverse sources: environmental monitoring, economic statistics, social surveys, geospatial observations, and customary knowledge systems. Each source has different quality characteristics, and accounts compilation must address those differences systematically through the principles set out in this Circular.

3.2 Six Dimensions of Data Quality

The SEEA EA identifies six dimensions of data quality relevant to environmental-economic accounting, plus the quality of the institutional environment in which the data are compiled (para. 2.86; see Section 3.3). The six dimensions apply directly to ocean accounts:

Definition: The degree to which the data serve to address the purposes for which they are sought by users.

Ocean accounts considerations:

• Do the accounts address the policy questions being asked?
• Are the spatial and temporal resolutions appropriate for the intended use?
• Are the ecosystem types and economic sectors adequately disaggregated?

Example: An ocean account designed to inform marine spatial planning needs fine spatial resolution, whilst one supporting national budget processes may require less geographic detail but more reliable monetary estimates.

Definition: The length of time between the data reference period and availability of the information.

Ocean accounts considerations:

• How current are the environmental monitoring data?
• Are there lags between economic activity and its recording in accounts?
• How frequently are accounts updated?

Example: Satellite-derived ecosystem extent data may be available within months, whilst large-scale economic surveys may have 1-2 year lags.

Definition: The degree to which the data correctly describe the phenomena they were designed to measure.

Ocean accounts considerations:

• What are the measurement errors in environmental monitoring?
• How reliable are sampling-based estimates of fish stocks or ecosystem condition?
• What are the error margins for economic survey data?

Example: Fish stock assessments may have wide confidence intervals. Accounts should report those intervals alongside the point estimate.

Definition: The degree to which data from different sources or compiled using different methods can be reliably combined and compared.

Ocean accounts considerations:

• Are classifications consistently applied across data sources?
• Are spatial boundaries aligned between environmental and economic data?
• Are accounting periods synchronised?
• Are changes in accounts over time attributable to genuine change rather than methodological or classification revisions? Breaks-in-series should be documented and, where feasible, back-casted (compilers should consult the relevant Section 3 accounts compilation circular for revisions practice specific to each account type).

Example: Economic data by industry may not align perfectly with spatially-defined ecosystem service flows, so bridging tables may be needed.

Definition: The availability of information to help users understand and properly use the data.

Ocean accounts considerations:

• Are methods and assumptions clearly documented?
• Are metadata standards applied?
• Are definitions consistent with international standards?

Example: Accounts should specify how services were measured and which ecosystem service classification is used, for example CICES V5.2 or the SEEA EA Reference List of Ecosystem Services (see SEEA EA Annex 6.1). For SEEA EA accounts, the SEEA EA Reference List is the primary classification, and CICES V5.2 may be used as a complementary mapping.

Definition: The ease with which users can access and use the data.

Ocean accounts considerations:

• Are accounts published in accessible formats?
• Are underlying data available for users who need more detail?
• Are there access restrictions (e.g., for confidential business data)?

Example: Accounts may be published in summary form with detailed underlying data available on request or through data portals.

Beyond the six dimensions of data quality, the SEEA EA emphasises the quality of the institutional environment in which data are compiled (para. 2.86), consistent with the UN Fundamental Principles of Official Statistics (Resolution 2013/21):

Table 3.6.1 sets out a recommended flagging system for ocean accounts tables.

Note: the “n/a” symbol replaces the double-hyphen form used in some earlier drafts. The double hyphen (--) is reserved for em-dashes in circular prose.

When account aggregates are derived from cells carrying different flags, the aggregate should carry the flag of the lowest-quality input (e.g., an aggregate of A and C cells carries flag C). Where the lowest-quality input contributes less than 5% of the aggregate value, compilers may exercise judgement and document the rationale.

3.7 Quality Improvement Strategies

Table 3.7.1 summarises four complementary strategies for progressive improvement of ocean accounts over time.

3.7.1 Global Consultation on the Technical Guidance

The quality improvement strategies above apply to the ocean accounts that countries compile. The GOAP Technical Guidance itself is subject to an equivalent quality assurance process: each circular is developed and published through a structured Circular Review and Global Consultation that applies the same principles of transparency, attribution, and documentation set out in Section 3.5 to the guidance documents. A circular passes through five phases over approximately eleven weeks (Table 3.7.2).

Four roles carry the process: the lead author (drafts and responds to comments), the review panel (reviews and summarises), a moderator (routes comments and escalates unresolved items), and the Secretariat (adjudicates, finalises, and publishes). The consultation is conducted openly. Comments are publicly visible while the window is open, and the full comment record is retained as a public archive (logged to an annex of the circular). Escalations are logged, and any rejected suggestion is accompanied by a documented justification. Draft circulars open for consultation are presented with an inline commenting overlay that lets readers select text and submit feedback directly on the page (see, for example, the TG-0.2 consultation page).

3.8 Tiered Implementation Frameworks

Throughout the GOAP Technical Guidance, tiered implementation frameworks are used to indicate graduated approaches calibrated to available data and institutional capacity. Tier designations are circular-specific, as each circular defines the tiers appropriate to its subject matter. In all cases Tier 1 represents the minimum viable approach, whilst higher tiers represent more rigorous methods requiring greater capacity or data availability. Compilers should select the tier consistent with their data availability and intended account use, and may progress to higher tiers as capacity develops.

3.9 Circular-Specific QA Guidance

Table 3.9.1 maps circulars in Sections 3 and 4 to the quality assurance focus of their topic. TG-4.1 to TG-4.6 each include a section on data quality or quality assurance. TG-3.1 to TG-3.5 and TG-4.7 have no dedicated quality assurance section (TG-3.5 discusses uncertainties and risks in its Section 5); for these topics, the principles in this Circular apply directly.

Authors: [To be confirmed]

Reviewers: [To be confirmed]

This circular draws on quality assurance frameworks from the SEEA EA (Chapter 2), the UN Fundamental Principles of Official Statistics, and the Generic Statistical Business Process Model (GSBPM).