Durability of Trained Judgment

Question. Does the trained effect persist after practice stops, how fast does it decay, and what is the minimal dose for lasting impact?

Analogy. Trained resistance could be one of two kinds of iron in a magnetic field. Soft iron is magnetic only while the field is held on it and goes dead the instant the field is removed; steel hardened past a certain point becomes a permanent magnet that holds the field on its own, with no power source. The study asks which one the trained effect is, how fast it fades once practice stops if it is the soft kind, and the smallest dose of practice that turns it from the soft kind into the permanent kind.

What's at stake. Every other study in the portfolio measures whether the training works while it is happening. None measures whether it survives the operator walking away. This is the question a defence sponsor asks of any intervention, does the capability persist or evaporate the moment the programme ends, and it is the question the business lives or dies on: an effect that fades is a subscription that churns, an effect that lasts is a capability that compounds. The proposal already bets the strong form of this in the onion model, that the capacity persists when the platform is denied, but it is asserted, not measured. There is also a sharp soft spot the durability question exposes. The training takes hold on the operator's own high-stakes problems, where the cost of being wrong is real. But the questions cognitive warfare most wants to bend are the opposite kind: low-personal-stake public ones, where being wrong about a political or collective issue costs the operator nothing directly, so the operator never bears down on them and never practises clear thinking there. A resilience that only holds during active practice is therefore switched off on exactly the questions the adversary targets, leaving the operator sharp on their own problems and defenceless on the ones that move a population; it reaches those questions only if the training has hardened into a standing habit that holds without active practice. So durability is not a nice-to-have but a precondition for the population-scale claim.

The two answers it decides between. Either the trained effect is specific learning that fades to baseline once practice stops, so resistance is a perishable state that needs the platform running to maintain; or it is a conditioned disposition that persists, carried by the operator's changed self-narrative and by a community whose norm keeps re-triggering it. The decay curve tells them apart: a steep fade toward baseline supports the first, a durable plateau supports the second.

What a null result would mean. If the effect decays to baseline within a short window after practice stops and no achievable dose produces a lasting plateau, the population-scale resilience claim and the onion-model tech-denial claim both need revision, because a capability that does not outlast active use cannot be the standing civil-defence asset the proposal describes. That is a finding about the theory, not the platform.

Why this matters to defence. Durability is the difference between a training course and a capability (DRDC Objective 6; and the personnel-development question under Objective 2). It changes a concrete decision: whether resilience training for personnel should be a one-time intervention or requires an ongoing refresher cadence, and what that cadence should be, which is exactly the minimal-dose output. The tech-denial-persistence question, whether trained judgment survives when the tool is gone, is the same question a NORAD, Arctic, or comms-denied planner asks.

How we would run it. Follow the same people over time: measure their resistance at the end of active practice, then again at intervals after they stop, compared against a group still practising and a no-platform control, using a wargame on identity-charged material. We vary the dose (missions completed, months of practice, number of two-way guide ties), so the result gives both a fade curve and the smallest dose at which the effect levels off instead of dropping. The still-practising group separates normal fading from the effect of stopping; the count of community ties tests whether peers re-triggering the habit is what keeps it alive. The only-us part is operators with genuine sustained practice and a real post-practice follow-up window, which a managed cohort cannot reproduce.

Earliest start. Stage 5: the study needs trained operators and, by construction, a post-practice follow-up window, so it starts with onboarded cohorts and runs longest of any study in the portfolio.