Open-Source Intelligence Methodology
The IPSC indexes and monthly digests are built on a signal-based approach to open-source intelligence collection. A signal is a discrete, source-verifiable event — a diplomatic statement, a military exercise, a regulatory decision, a reported incident — assessed by an analyst as relevant to a defined topic area.
How signals are gathered
Signals are collected by IPSC's network of contributing research centres, each assigned to a specific analytical domain: China, India, the Philippines, Japan, Cross-Strait dynamics, the Indo-Pacific region, maritime corridors, cyber and technology, and the United States. Each centre works from domain-specific collection briefs that define the topics, actors, and event types that fall within scope.
Collection is entirely open-source. Signals draw on publicly available material — government statements, official documents, diplomatic communiqués, defence announcements, credible media reporting, and published data. No classified, proprietary, or commercially restricted intelligence is incorporated.
Collection is conducted by both human analysts and automated processes working from the same briefs. Human analysts apply judgement in identifying, extracting, and assessing signals; automated processes extend coverage across high-volume periods.
How signals are assessed
Each signal is assessed along two dimensions: magnitude and direction.
Magnitude reflects the assessed significance of the event — its scale, the seniority of actors involved, and the degree of change from prior conditions. Magnitude is coded against structured anchors defined for each domain, so that equivalent events are rated consistently across the series.
Direction reflects whether the event is assessed as escalatory or de-escalatory relative to the relevant security or policy baseline.
Signals are also tagged with primary actors, geographic location, and an indicator name — a standardised label describing the type of event recorded.
How scores are computed
Monthly index scores are computed by aggregating the assessed magnitudes of signals collected within a given period. A normalisation step ensures that events generating multiple signal records — an announcement followed by a reported action, for instance — are weighted as a single event rather than inflating the score.
Scores are best read as a relative measure of signal activity over time within a given domain or corridor. They are not normalised against collection volume, and cross-domain comparisons are not supported: collection scope differs across centres, and the index is not calibrated to produce comparable absolute scores between them.
Monthly analytical outputs
Monthly digests synthesise the signal record for each research centre into a structured analytical output. They identify the dominant signals for the period, interpret patterns in the data, and situate the month's findings in the context of the preceding series.
Digests draw on the same underlying signal corpus as the indexes and serve as the narrative layer alongside the quantitative scores.
The IPSC indexes measure signal activity — what analysts observed and how it was assessed. They do not assess the intent behind the events recorded, produce probability estimates for future events, or incorporate judgements about outcomes. Index scores are an evidential input for further analysis, not a risk rating or forecast.