EDC

Research outputs

Research manuscripts, reference materials, and tools.

Selected work from my Collective-State Inference research program: what each contribution offers, its current status, and where to explore it.

Featured manuscript
In peer reviewTheoretical research manuscript · Edward D. Clark

When does evidence about interacting participants justify an AI system's claim about the condition of a collective? This manuscript proposes the theoretical requirements for that interpretation, including collective boundaries, how evidence combines, uncertainty, and comparison with simpler explanations.

Status: Unpublished and currently in external journal peer review following editorial suitability assessment. It has not been accepted, and CSI has not yet been empirically validated.

View the manuscript status record Understand the research question
Archived reference
Archived and citableConceptual reference · Edition 0.1 · 2026

CSI Reference, Edition 0.1

A fixed edition of the program's conceptual foundations, including ten reference articles, construct boundaries, methodological principles, scholarly sources, and editorial policies. The archive provides a persistent record alongside the evolving online Reference.

Archival publication establishes citability and traceability; it does not establish peer review or empirical validation.

Clark, E. D. (2026). CSI Reference, Edition 0.1. Collective-State Inference Research Program. https://doi.org/10.5281/zenodo.21701697

Review the archival release Open the DOI record
Conceptual illustration
Synthetic illustrationPublicly available

CSI in a collaborative listening session

A worked example follows a hypothetical listening group from observable participation to a candidate estimate of collective engagement. It shows why the distribution, reciprocity, and persistence of interaction may matter alongside activity totals.

The data are entirely synthetic. This is a conceptual illustration, not an implemented or empirically validated CSI system.

Explore the listening-session example Read the classroom example
Research infrastructure

Tools supporting the research process.

Private prototype · v0.1

CSI MCP

An experimental machine-readable interface for retrieving approved CSI resources, checking proposed observation sets, and constructing traceable evidence ledgers. It supports preparation for research; it does not estimate or score collective states.

Availability: Public overview; private development and evaluation. The endpoint and source repository are not publicly available.

Read the capabilities and access boundaries
Exploratory instrumentation

CSI Research Observatory

A public prototype showing how delayed, aggregate website interaction observations become audience-pattern indicators and candidate evidence.

Website visitors are not assumed to form a qualified bounded collective. The Observatory does not publish validated CSI collective-state estimates.

Explore the public Observatory
Work in progress

Methods and studies in development.

Operationalization protocol Internal draft · Under methodological review
Develops construct definitions, indicators, observation windows, and comparison criteria for collective engagement. It is not yet a published, preregistered, or validated research output.
Empirical comparison and validation Planned studies
Will test candidate representations against simpler baselines and examine validity, calibration, and generalization. These studies have not yet established empirical results.
Follow the detailed research program
Technical education

Practical instruction.

Through How’d You Code That, I share instruction in software engineering, cloud architecture, and application development. These educational contributions are separate from the scholarly research record.

Browse selected videos