RESEARCH
Research
How Intelligence Works studies how intelligence is assessed, supported, and used in real-world contexts.
Psychometric research provides established methods for measuring cognitive abilities and understanding individual differences. HIW builds on this foundation by studying how abilities interact with health, state, tasks, environments, support, institutional demands and changing technologies.
HIW is developing a research programme around one central question: how do these factors come together to shape performance across real-world contexts, and how can that knowledge help people use their cognitive abilities more effectively?
What Is Already Established
The project begins from established intelligence science. Cognitive abilities can be measured. Cognitive tasks tend to correlate. General cognitive ability is a useful construct. Specific profiles can matter. Test scores require validity, reliability, fairness, and interpretation. Cognitive measures can predict important outcomes at the population level. Context, health, development, opportunity, and support shape performance.
HIW does not need to reinvent that foundation. It builds from it.
What HIW Is Asking
HIW focuses on the pathway from capacity to performance.
It asks what happens between a person’s cognitive abilities and the final observable outcome: task demands, environmental conditions, current state, access, support, tools, expression, and institutional judgement.
This is a research direction, not a final settled framework.
Current Research Layers
The research programme includes current operationalization using existing measures, the Functional Intelligence Framework as a developing systems architecture, the Open Intelligence Measurement Instruments initiative as a staged measurement programme, prospective predictions and validation plans, public research records through OSF and Zenodo where appropriate, and restricted materials where openness would compromise validity, privacy, test security, or ethical use.
Scientific Status
The website should label research claims clearly: established evidence, supported interpretation, HIW proposal, active manuscript, preprint, peer-reviewed publication, public research record, or restricted material.
This is not hedging. It is how the project earns trust.
Research Questions
How Intelligence Works is organised around one central question: how can intelligence be assessed in the real world, and how can people be better supported in using their cognitive abilities?
That question connects intelligence science to daily life, education, work, health, relationships, institutions, technology, and support.
Core Questions
What does a real-world task actually require? How do environments change what people can do? Why does the same person perform differently across time and context? How do state, fatigue, stress, illness, disability, and access change performance? How should support be evaluated? How can intelligence be assessed without reducing a person to final output? How can open science coexist with test security and responsible administration?
What Would Count As Progress
Progress would mean better task descriptions, better environmental records, better support models, better interpretation of assessment, better public research records, and better ways to distinguish capacity from the systems that help or block its expression.
Active Research
How Intelligence Works includes active research that is still developing.
That work should be read as provisional. A framework under review is not a consensus. A proposed taxonomy is not a settled ontology. A research programme is not validated simply because it is coherent.
Functional Intelligence Framework
The Functional Intelligence Framework is a developing systems architecture for distinguishing intelligence from performance.
It asks what intervenes between cognitive capacity and observable behavior: state, access, task demands, environment, support, output conditions, and performance judgement.
The framework is being developed through manuscripts, appendices, public-dataset evaluation, prospective predictions, and future validation work.
Open Intelligence Measurement Instruments
OIMI is a staged measurement programme for developing, validating, documenting, and governing future instruments.
It is not a completed public test battery. Inclusion in a registry does not mean an instrument is validated, unrestricted, or ready for operational use.
OIMI distinguishes between concept definition, prototype development, technical pilot testing, formal validation, independent replication, operational release, and archival status.
Publications
This page tracks public outputs from How Intelligence Works.
It may include peer-reviewed articles, preprints, OSF records, Zenodo records, public appendices, protocols, prediction registries, codebooks, non-sensitive datasets, and framework summaries.
Each output should be labelled by status: submitted manuscript, preprint, peer-reviewed article, public appendix, protocol, registry excerpt, restricted material, or superseded version.
Publication makes a work citable. It does not make every claim settled.
Research Infrastructure
How Intelligence Works is building research infrastructure for studying intelligence in real-world contexts.
The goal is not only to explain intelligence. It is to create better ways to study how intelligence is assessed, supported, and used across health, education, work, daily life, institutions, and AI-supported environments.
The Infrastructure Problem
Much of intelligence research depends on specialized tools, proprietary assessments, restricted manuals, protected test content, and expert interpretation.
Some restriction is necessary. If test items, scoring keys, or sensitive protocols are exposed without control, validity and ethical use can be damaged.
At the same time, intelligence is too important to be understood only by people and institutions who can afford access to proprietary tools.
HIW works toward a careful middle path: open enough to support public knowledge, research collaboration, and accountability; protected enough to preserve test security, privacy, validity, and responsible use.
OIMI
The Open Intelligence Measurement Instruments initiative is a prospective measurement infrastructure for the Functional Intelligence Framework.
OIMI aims to support the development of independently authored instruments that can be documented, validated, governed, and, where appropriate, made available for public-interest research and education.
OIMI is not a completed public test battery. It is a staged programme for moving instruments through concept definition, prototype development, technical pilot testing, scientific validation, replication, operational use, or archival status.
Open documentation does not mean unrestricted administration. Some instruments may require supervision, professional qualification, restricted access, or delayed release.
How The Infrastructure Fits Together
The Human Task Atlas begins with the task: what the person is actually being asked to understand, remember, decide, coordinate, or do.
The Environmental Registry adds the conditions: noise, time pressure, social expectations, interface design, fatigue, support, institutional rules, and other factors that shape whether ability becomes performance.
The Intelligence Observatory follows the person over time, showing how profile, state, support, and context change what is possible.
The Performance Lab brings those layers together in repeated performance episodes, testing what changes when the task, support, tools, or environment changes.
OIMI provides a staged pathway for developing future measurement instruments, while OSF and Zenodo provide public research records where appropriate.
NEXT
Help build the evidence.
The research develops through open questions, shared infrastructure, criticism and collaboration.