Control Condition Fidelity Checklist¶
Verification checklist — instantiates Control-Condition Specification
An item-by-item verification that the control arm, as actually delivered, matched its specification — that the intended differences were present and the required equivalences held.
A control arm can be specified in exhaustive detail and still be delivered wrong: a provider slips active advice into the "no-advice" arm, a site skips the matched follow-up call, an incentive meant to be equal is quietly larger in one group. Control Condition Fidelity Checklist is the verification instrument that closes that gap. It takes the two things the specification promised — the features that were supposed to differ between treatment and control, and the features that were supposed to be equal — and checks, item by item, session by session, whether the delivered control honored them. It does not define the control and it does not watch continuously for leaks; it confirms conformance at defined verification points, producing a pass/fail record of whether the intended contrast was actually enacted.
Example¶
A trial of computerized cognitive behavioral therapy for depression uses a "psychoeducation-only" control: the same online platform, the same login schedule, but without the interactive skill-building modules that are the treatment's active ingredient. On paper the arms differ only in those modules and match on everything else — platform, dosage schedule, therapist check-in length.
To make sure that is what happened, the team runs a Control Condition Fidelity Checklist against a sample of control-arm sessions and check-ins. The contrast items ask: was the active skill-building module genuinely absent? Did the therapist avoid coaching skills during the check-in? The equivalence items ask: was the check-in the same length as in the treatment arm? Were the same symptom measures administered on the same days? Reviewing recordings, the auditor finds that in one therapist's control sessions, brief coping-skills coaching crept in — a contrast-item failure — and flags those cases. Because the checklist caught it, the team can quantify the drift, retrain the therapist, and, in analysis, run a sensitivity check excluding the affected participants. Fidelity, here, is a measured quantity rather than an assumption.[n1]
How it works¶
The checklist is derived directly from the control-arm specification and split along a single, clarifying axis:
- Turn the contrast map into check items. For every feature the design intended to differ, write a verifiable item ("active component absent in control") that an auditor can score as present/absent.
- Turn the equivalence boundary into check items. For every feature that must be held equal across arms (contact length, measurement schedule, incentives), write an item confirming the control matched the treatment.
- Score at defined verification points. Apply the checklist to sampled sessions, visits, or records at set intervals, producing a fidelity rate rather than a one-off impression.
- Route failures to a response. Each failed item triggers a pre-agreed action — retraining, exclusion, or a documented deviation — so the check has teeth.
Tuning parameters¶
- Sampling fraction — how many control encounters are audited. Higher fractions catch rare drift but cost reviewer time; low fractions risk missing localized failures.
- Item resolution — how finely each "differ" and "hold equal" requirement is decomposed. Fine items catch subtle infidelity but lengthen the checklist and invite rater fatigue.
- Pass threshold — how strict the conformance bar is before an encounter counts as faithful. A strict bar protects the contrast; a lax one tolerates noise but flags fewer false problems.
- Verification cadence — whether checks are one-time, periodic, or triggered by suspicion. More frequent checks catch drift early at higher operational cost.
When it helps, and when it misleads¶
Its strength is that it converts fidelity from faith into evidence: it tells you whether the elegant control you designed is the control you ran, and it does so in a way that distinguishes the two kinds of failure — a lost intended difference versus a broken required equivalence. That distinction is what lets a team fix the right thing.
It misleads when the checklist is mistaken for completeness. A checklist only verifies what its items name; a subtle form of infidelity nobody thought to list passes silently, and a high fidelity score can then certify a contrast that was actually compromised in an unlisted way. It also verifies at points, so drift between checks or slow contamination across the whole sample can escape it — which is why a checklist is not a substitute for continuous surveillance. The guarding discipline is to derive items straight from the specification (not from convenience), to keep a free-text "other deviation" capture alongside the fixed items, and to pair the checklist with a contamination log for the between-check gaps.
How it implements the components¶
treatment_control_contrast_map— it verifies the "what should differ" half: each intended difference between arms becomes an item confirming the active ingredient stayed out of the control.comparator_equivalence_boundary— it verifies the "what should be equal" half: each matched feature becomes an item confirming the control did not drift off the equivalence boundary.
It does not author the specification it checks (control_condition_definition, outcome_and_timing_alignment) — that is its nearest twin, Control Arm Protocol, which declares intent where this checklist confirms delivery — and it does not maintain the running breach ledger over the whole study (control_condition_integrity_guardrail, contrast_bounded_reporting_record), which is Contamination Monitoring Log.
Related¶
- Instantiates: Control-Condition Specification — it verifies that the specified contrast was the contrast actually delivered.
- Consumes: Control Arm Protocol supplies the specification whose "differ" and "hold equal" requirements become the checklist's items.
- Sibling mechanisms: Control Arm Protocol · Contamination Monitoring Log · Attention Control Script · Placebo or Sham Procedure · Standard-Care Comparator Specification · Usual-Care Inventory Form · Waitlist Control Schedule · External Control Justification Memo
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: An item-by-item verification that the control arm, as actually delivered, matched its specification — that the intended differences were present and the required equivalences held, making its operative form a bounded evaluation of existing evidence or work that produces a finding or disposition.
Independent corroboration: The frozen evidence defines Control Condition Fidelity Checklist as 'An item-by-item verification that the control arm, as actually delivered, matched its specification — that the intended differences were present and the required equivalences held', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Statistics & Experimental Design
Origin pattern: Single lineage
Present-day reach: Specialized
Rationale: Experimental fidelity methodology cohered itemized verification that the delivered comparator preserved intended differences and required equivalences.
Related originating lineages:
- Medicine & Healthcare — Clinical-trial monitoring institutionalized session- and site-level fidelity checks against a control-arm protocol.
Review resolution: Experimental fidelity assessment is the coherent method lineage; clinical-trial monitoring is a genuine specialized institutional lineage for control-arm checks.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Treatment (or implementation) fidelity is the degree to which an intervention — or, here, a control condition — is delivered as designed. Measuring control-arm fidelity is as important as measuring treatment fidelity: a contaminated control biases the effect just as surely as a poorly-delivered treatment does. ↩