How SiPAT measures performance
Methodology
A transparent guide to the data sources, calculations, and limitations behind SiPAT's transparency and accountability scores.
Purpose
This methodology establishes the principles, data sources, processing procedures, classification methods, and scoring mechanisms used by SiPAT to transform government records into structured transparency and accountability indicators. It provides users with a clear understanding of how information presented on the platform is collected, processed, organized, and interpreted.
What SiPAT Measures
Transparency concerns the availability and consistency of government disclosures and transparency-related records. SiPAT uses Full Disclosure Policy Portal (FDPP) records and Transparency Seal documents to examine whether government entities have published the required information and documentation within the applicable reporting periods. Accountability concerns the extent to which government entities address issues identified through Commission on Audit (COA) reports. SiPAT examines documented audit findings and the recorded resolution of COA recommendations to provide an indicator of recommendation implementation.
Interpretation and Limitations
SiPAT results are intended to help users examine documented government records, identify patterns, and determine areas for further review. They may be used to assess public disclosures, audit findings, implementation of audit recommendations, and available indicators across government entities and reporting periods. However, SiPAT scores should not be interpreted as definitive assessments of an agency’s overall performance, integrity, efficiency, legality, or absence of irregularities.
Scope and Coverage
SiPAT covers government entities across three primary accountability groups: National Government Agencies (NGAs), Local Government Units (LGUs), and Government-Owned and Controlled Corporations (GOCCs). These agencies are organized within the platform to allow users to access their available audit reports, transparency records, findings, and corresponding performance indicators. The availability of records may vary by agency and reporting period. When sufficient data is not available for a particular agency or period, SiPAT does not assign a numerical score and instead indicates the absence of available data.
Data Sources
SiPAT consolidates publicly available government records from selected official sources. Commission on Audit (COA) Annual Audit Reports serve as the primary source for audit findings, recommendations, and accountability information, with records linked to their original reports for traceability. The Full Disclosure Policy Portal (FDPP) provides transparency and financial disclosure records for Local Government Units, which SiPAT uses exclusively for transparency measurement and the Transparency Score. The LGU Portal serves as a separate source of LGU-related documents maintained within SiPAT’s searchable document repository. These sources are kept distinct according to their purpose: COA records support audit and accountability analysis, FDPP records support transparency measurement, and LGU Portal records provide additional government documentation.
Data Collection and Processing
SiPAT collects source documents in ZIP, PDF, DOC, DOCX, XLS, and XLSX formats, associating each with its originating agency, source, year, and document type before storing them through a structured pipeline that separates raw files from processed information (documents → document_files → file_contents → document_chunks → document_extractions → findings); once ingested, documents undergo text extraction and are divided into token-counted chunks, which are then passed through AI-assisted extraction to identify and structure individual findings — including title, description, category, plain-language explanation, and recommended action.
Severity Classification
Each audit finding extracted from a COA Annual Audit Report is assigned a severity level of Low, Moderate, or High. Severity is calculated as a weighted score across four risk dimensions — Impact, Financial Exposure, Control Weakness, and Recurrence/Urgency. Reflecting SiPAT's own risk-prioritization framework, informed by general audit risk principles. Financial Exposure is derived directly from monetary amounts stated in the finding, while Impact, Control Weakness, and Recurrence/Urgency are estimated through AI-assisted analysis of the finding's descriptive language. Severity is an interpretive prioritization tool intended to help users identify findings that may warrant closer attention, it does not indicate legal culpability or confirmed wrongdoing.
SeverityScore = 0.35(Impact) + 0.30(FinancialExposure) + 0.20(ControlWeakness) + 0.15(RecurrenceUrgency)
| Score Range | Severity Level |
|---|---|
| 0.0 – 2.9 | Low |
| 3.0 – 3.9 | Moderate |
| 4.0 – 5.0 | High |
- Impact (35%)
- Effect on public service, compliance, or public trust. AI-estimated from the finding's description; scored 1–5
- Financial Exposure (30%)
- Monetary value at risk, as stated in the audit finding; scored 1–5 based on defined peso-value ranges
- Control Weakness (20%)
- Strength or absence of internal controls related to the issue. AI-estimated from the finding's description; scored 1–5
- Recurrence/Urgency (15%)
- Whether the issue is isolated, recurring, or time-sensitive. AI-estimated from the finding's description; scored 1–5
Transparency Score
For LGUs, the Transparency Score is calculated from Full Disclosure Policy Portal (FDPP) submissions, as described above. For National Government Agencies (NGAs) and Government-Owned and Controlled Corporations (GOCCs), which are not covered by the FDPP, the score instead uses each agency's Transparency Seal compliance, the set of documents (e.g., agency mandate, budget, financial reports, procurement contracts, annual reports) required to be posted on their official website under DBM's Transparency Seal policy. The same cumulative logic applies: documents required once per year carry forward once posted, while documents required repeatedly are counted separately for each required period. Since the required document sets and posting frequency differ between FDPP and the Transparency Seal, the score is computed the same way structurally but against a source-appropriate checklist and denominator for each entity type.
Transparency Score = (Documents Posted ÷ Documents Required) × 100
- Documents Posted
- Number of required documents confirmed submitted or published
- Documents Required
- Total documents required based on entity type (FDPP checklist for LGUs, Transparency Seal checklist for NGAs/GOCCs)
Accountability Score
For LGUs, NGAs, and GOCCs alike, the Accountability Score is calculated from Commission on Audit (COA) Annual Audit Reports, specifically the recommendations issued and their recorded resolution status. Unlike the Transparency Score, the data source does not vary by entity type — all three accountability groups are audited by COA under the same reporting framework, so the same recommendation set and weighting scheme (Implemented, Partially Implemented, Other/Not Implemented) applies uniformly. What does vary is the reporting period used: the score is calculated based on each agency's latest available COA report year, since agencies are not always audited or their reports released on the same annual cycle. As a result, the score reflects recorded implementation as of the most recent audit data available for that specific agency, rather than a fixed calendar year applied uniformly across all entities.
Accountability Score = (Total Implementation Value ÷ Total Recommendations) × 100
- Total Implementation Value
- Sum of implementation values across all recommendations. Implemented = 1.0, Partially Implemented = 0.5, Other/Not Implemented = 0
- Total Recommendations
- Total number of recommendations recorded for the agency's latest available COA report year
Responsible Interpretation
SiPAT scores should be interpreted as indicators derived from available government records and should not be treated as definitive assessments of an agency's overall performance, integrity, or legality. A high Transparency Score does not necessarily mean an agency has no audit findings, as transparency and accountability are measured independently. Likewise, a low Transparency Score does not by itself establish wrongdoing; it may reflect incomplete disclosure rather than misconduct. An Accountability Score reflects only the recorded implementation of recommendations identified in an agency's latest available COA report, not a comprehensive evaluation of the agency's overall conduct. Users are encouraged to consult the underlying source documents referenced within SiPAT for complete context before drawing conclusions about any individual agency.
Read with context
Scores are indicators, not verdicts. They summarize the records available to SiPAT at a point in time and should be read alongside the original COA and FDPP documents.