The Institute of Agency Science

Agency, made measurable.

Lives are decided in half-second moments. The Institute of Agency Science exists to measure the human capacity to choose well in them, and to make that capacity teachable, so people can be trained to make better decisions when it matters most.

The Institute of Agency Science is an independent research institute and standards body being established for one purpose: to make human agency measurable, teachable, and protected as a global standard. The skill of choosing well in those moments has never been measured in its own right. The Institute is being built to measure it, to govern the measure, and to give the science away.

Read the Lexicon The 2036 Bet

The half-second problem

A life is steered by moments that last less than a second. The hand that reaches for the phone or does not. The sentence spoken in anger or held. The task begun or deferred one more day. These moments decide health, money, marriages, and careers, and they arrive by the hundred, every day, for every person alive.

Medicine learned to measure the body, and lifespans doubled. Psychometrics learned to measure intelligence, and education reorganized around it. The skill of choosing well inside the half-second has never been measured in its own right: the field measures its neighbors, grit, self-control, conscientiousness, executive function, but the choosing itself, at the moment it happens, has no instrument, no shared vocabulary, and no standard for claims about it.

That gap is the Institute of Agency Science's founding problem.

The claim is documented, instrument by instrument: the prior-instruments annex names every adjacent measure, states fairly what it captures, and records the precise sense in which the choosing itself remained unmeasured.

A timeline of a single decision moment: a cue arrives, a default response starts to run, and inside roughly half a second there is a window where a person can choose. The neighboring constructs the field measures sit around the window; the choosing inside it is marked as never measured in its own right. — Figure 1. The half-second problem. The field measures the neighbors of the moment; the choosing inside it has never been measured in its own right.

The Institute holds a three-part mandate, tied to one north star: make agency the literacy of the AI age.

Teachable

Define agency precisely enough that it can be taught: an open construct, openly documented, free to cite, so the skill of choosing well can be trained deliberately rather than left to chance.

Measurable

Build and steward an open instrument for measuring agency, with its limits stated as plainly as its readings.

Protected

Govern how the measure may and may not be used, so a measure meant to serve a person is never turned against one.

What a measure makes possible

Measurement is how a human capacity stops being a matter of opinion. Once agency can be measured, it can be studied. Once it can be studied, it can be taught. Once it can be taught, it can be trained deliberately: in schools, in clinics, in households, on an open standard that belongs to everyone. The half-second that decides a life stops being invisible.

That is the Institute's wager: a civilization that can measure the skill of choosing well can raise it. The wager is stated as a wager, and it is held to the evidence discipline on this page. No claim the Institute makes may exceed what the published evidence has earned.

Continue: The Lexicon · Working papers

The instrument

The Agency Quotient (AQ). [ Evidence tier: [hypothesis-grade] ]

The Agency Quotient (AQ) is the Institute's instrument for measuring the trained capacity to choose and execute the hard, right action at decision moments. It examines four capacities exercised at moments of choice: to see the moment, to start what one has chosen, to stop a pulled response, and to steer toward the chosen outcome, assessed across the domains of a life and reported as a profile of capacities, not a single number. It is an instrument, not a predictor.

No such instrument has existed before, and building one honestly means stating exactly where it stands. AQ is built and documented; its reliability and predictive validity are not yet established. That is what the [hypothesis-grade] badge means, and the Institute wears the badge in public. Validation is planned to be independent and pre-registered, and the Institute will publish those results whether or not they favor the instrument.

Continue: The AQ Standard

We wrote the bar down before the results came in

Most new measures announce their standards after the data arrives. The Institute fixed its bar first, in public. A new measure of agency earns the name only if it shows incremental explanatory value beyond grit, locus of control, self-efficacy, conscientiousness, self-regulation, and executive function. The comparators are named in advance, so the test is binding whichever way it falls. AQ has not yet been shown to clear that bar, and clearing it in the open is the program.

Every claim the Institute makes climbs a public evidence-tier ladder: [hypothesis-grade], [pilot-validated], [panel-validated], [institutional]. Each rung is bought only with published evidence, and no claim may borrow a rung it has not earned.

The same discipline runs a decade deep. The 2036 Bet is the Institute's single public, dated, ten-year wager, scored openly every year, toward or away, including in the years the evidence runs against it. An institution willing to be wrong in public, on the record, on a schedule: that is what a standards body is.

Continue: The 2036 Bet · Where validation stands

The evidence-tier ladder for AQ, running from hypothesis-grade, where it stands today, through pilot-validated and panel-validated to institutional, with each rung earned only by published evidence. — Figure 2. The evidence-tier ladder. AQ stands at hypothesis-grade; each higher rung is bought only with published evidence, never with prose.

Open by default

The Human Genome Project placed the genome in the public domain, and a generation of science grew on top of it. The Institute takes the same posture toward the science of agency. The Lexicon, the Charter, the Agency Bill of Rights, and the working papers are intended for release under CC-BY 4.0 at launch: free to read, cite, quote, translate, and build on. The definitions are free; the marks are reserved, so a label keeps meaning what it says.

A standard that cannot be read and quoted in full is not a standard. Openness is how a measure becomes a shared reference rather than a private asset.

Continue: Read the Lexicon

The open-by-default pipeline: the bar and the comparators are fixed in public first, the validation study runs pre-registered, the result is published whether favorable, null, or negative, and the record stays open for anyone to cite and test against. — Figure 3. The open-by-default pipeline. The bar is fixed before the data, and the result is published whichever way it falls.

Built to be governed, not sold

A measure of human capability is powerful enough to be abused, so the Institute is being built with the protections in the architecture rather than in the fine print. The person holds the score; the Institute holds the instrument. The architecture is designed with no central lookup bureau: no place where an employer, lender, insurer, or school can pull a person's score by name. The only third-party read is a share token the person issues, can revoke, and can watch being used. This describes the architecture as designed, not a pledge about every possible future.

The Agency Bill of Rights is the enumerated protection layer for anyone touched by agency measurement or training. Today, AQ may not be used as a selection gate for hiring, lending, school admission, insurance, custody, or care access, and the architecture makes those uses inexpressible at the instrument itself. This is stated as a current discipline: AQ is not yet validated for job-relatedness or adverse impact, so it is not licensed for selection.

Independence is a design requirement with a public test: judge the Institute by whether it ever rules against its own founder. The founding instruments provide the authority to publish unfavorable findings about any method including the founder's, to refuse abusive uses by any licensee, to decertify any practitioner, and to set certification criteria without commercial veto.

Continue: The Agency Bill of Rights · Governance

The Institute in full

Lexicon, the open vocabulary of agency science, free to cite.
The Standard, the AQ methodology, published with its limits.
Working Papers, the open research record.
The 2036 Bet, the ten-year public wager, scored annually.
Charter, the founding governance document.
Bill of Rights, the protection layer, free to adopt.
Certification, CAE credential governance.
Governance, independence, the board, and formation status.
About, what the Institute is.
Support, fund the mission.

Support the Institute

The Institute is funded as a mission, not sold as a product. There is nothing to buy on this site. A gift funds the creation of a public scientific asset: the validation science, the open standard, the credential governance, and the published research that a neutral standards body exists to steward. The Institute is being established as a 501(c)(3); tax-deductibility depends on that filing.

Continue: Fund the Institute