Free-Recall-Then-Recognition Probe¶

Diagnostic estimation — instantiates Encoding–Retrieval Context Alignment

Asks first for unaided recall, then for recognition, and reads the gap between them to tell 'never stored' apart from 'stored but not retrievable.'

The Free-Recall-Then-Recognition Probe is a two-stage diagnostic. Stage one asks for free recall — unaided, under the real demand. Stage two asks for recognition — the same material cued, offered in a lineup, or multiple-choice. The pattern across the two stages diagnoses the failure: fail both and the material was never stored (an availability problem — re-encode it); fail free recall but pass recognition and the material is stored yet inaccessible (an accessibility problem — fix the cues). Its defining move is separating availability from accessibility by contrasting two retrieval demands on the same trace — a distinction no single test can draw.

Example¶

An eyewitness cannot spontaneously describe a suspect — free recall comes up nearly empty. Minutes later she picks him out of a fair lineup quickly and confidently — recognition succeeds. The probe reads this split cleanly: the memory is there, but it wasn't accessible to open-ended recall, so the remedy is better retrieval cues — a structured, cue-rich interview — not re-exposure to the event. Had both stages failed, no amount of cueing would help, because nothing was encoded to begin with. The probe's value is telling the investigator which world they are in before they pour effort into the wrong remedy.^[1]

How it works¶

You first pin down the real retrieval demand — usually unaided and generative — and administer free recall at that level. Order matters: free recall comes strictly before recognition, so the recognition options don't contaminate the unaided attempt. Then you administer recognition and read the resulting two-by-two. What distinguishes this probe is its axis — demand level (free versus recognition) rather than which cue matters or which rival interferes — and that its verdict directly triggers the right maintenance: re-encode when the trace is unavailable, realign the cues when it is merely inaccessible.

Tuning parameters¶

Order and contamination control — free recall strictly before recognition, and how much delay sits between them.
Recognition difficulty — how plausible the foils are; too-easy recognition inflates the "stored" verdict.
Stage-one demand realism — how closely the free-recall task matches the true, unaided use demand.
Pass threshold — how much partial recall counts as success at each stage.
Re-test cadence — one-shot, or repeated over time to catch decay that only shows up later.

When it helps, and when it misleads¶

Its strength is routing remediation: it stops teams from re-teaching material that is actually a cueing problem, and from bolting cues onto material that was never learned. It misleads when recognition is made too easy — everything then looks "stored," because recognition flatters — or when the stage-one demand doesn't match reality. The classic misuse is reporting the recognition score as mastery ("they scored 90%") when the real task demands unaided recall: the illusion of knowing, dressed up as evidence. The guard is to match stage-one demand to genuine use, keep recognition honest with plausible foils, and act on the pattern — trigger re-encoding versus re-cueing — rather than on the headline number.

How it implements the components¶

The probe realizes the diagnose-and-route side of the archetype — measuring the nature of the failure and pointing to the fix:

retrieval_demand_definition — it pins down the real retrieval demand (unaided versus cued) the material must meet, and tests explicitly at that level rather than an easier one.
maintenance_and_re_encoding_trigger — the availability-versus-accessibility verdict triggers the correct maintenance action: re-encode when nothing is stored, realign cues when the trace is merely inaccessible.

It does not isolate each cue's individual contribution (diagnostic_cue_set, from Cue-Diagnosticity Ablation Test) or emit an interference/discrimination signal against competitors (retrieval_outcome_signal, from Interleaved Competitor Retrieval Test); it separates storage from access.

Instantiates: Encoding–Retrieval Context Alignment — the probe supplies the storage-versus-access diagnosis that tells the alignment which remedy to reach for.
Consumes: nothing required; its verdict feeds re-encoding work such as Post-Event Re-Encoding Debrief and cue redesign.
Sibling mechanisms: Cue-Diagnosticity Ablation Test · Interleaved Competitor Retrieval Test · Post-Event Re-Encoding Debrief · Scenario-Based Retrieval Test · Representative-Environment Simulation · Spaced Retrieval Scheduler · Context Reinstatement Protocol · Context Translation Card · Context-Switch Recall Drill · Cue-Fading Schedule · Environmental Retrieval Cue · External Checklist or Job Aid · Mnemonic Cue Pairing · Transfer-Appropriate Processing Rehearsal · Varied-Context Retrieval Practice

References¶

[1] The availability–accessibility distinction — a trace can be present in memory (available) yet not reachable with the cues at hand (inaccessible); recognition can succeed where free recall fails (Tulving & Pearlstone). The probe operationalizes exactly this contrast. ↩