To investigate low IOA in ABA data, freeze both source records and check matched exposure, response definitions, timing, access, visibility, training, observer independence, event matching, and recording systems. Classify each disagreement before selecting a repair. Low agreement identifies uncertainty in the sampled measurement process. It does not identify which observer is correct or whether the client's behavior changed.

Freeze the evidence

Lock original records, timestamps, procedure version, observer assignment, access logs, and calculation before discussion.

Classify discrepancies

Separate occurrence, category, prompt, timing, duration, trial validity, event matching, missingness, and exposure problems.

Check system conditions

Review clock drift, device lag, form logic, visibility, workload, and interruptions.

Match repair to cause

Use definition revision, calibration, access repair, system correction, resampling, or clinical review according to the evidence.

Start with a discrepancy ledger

Create one row for every eligible unit. Preserve each observer's original score, exposure state, timestamp, response opportunity, prompt timing, communication mode, and system version. Add a disagreement category and evidence source without changing the raw entries. This ledger turns a low percentage into a set of testable process questions.

Review observers separately before a joint debrief when score sharing could influence recall. Ask each person how the definition was applied and which units felt uncertain. Compare those accounts with timestamps, video or system evidence when use is authorized, and the person's own accessible report.

Reconcile Imani's five disagreements

Seven agreements across twelve opportunities leave five disagreements. The example accounts for all five: three involve prompt timing and two involve whether an AAC selection met the response definition. The categories are mutually exclusive in this fictional ledger. If a real unit fits more than one cause, keep a primary category plus secondary flags so category totals are not mistaken for unique counts.

The prompt-timing cluster suggests checking the window boundary, clocks, and whether the prompt began before the response was complete. The AAC cluster calls for review of device access, the recognized response form, partner wait time, and whether observers could see the selection. Preserve Imani's communication as client evidence instead of treating observer agreement as the authority on what Imani meant.

Test the repair with a new sample

Version the clarified definition and training material. Use examples and nonexamples, then confirm that observers can apply the rule independently before returning to live observation. Repair AAC visibility or access first. Schedule a representative sample with separate records and the same prospectively defined denominator.

Keep the original 7 of 12 result in the record. The new sample measures performance under the revised process and should be reported as a separate version. Improvement after several changes supports operational follow-up but does not isolate which change caused it.

Apply the field control to one defined question

The low-IOA investigation log for Imani begins with the exact decision, the responsible qualified role, and the evidence needed. Preserve original observer records, matched exposure, the selected calculation, raw numerator and denominator, unrounded result, and every excluded or unresolved unit.

Work the example for Imani

Imani's observers agree on 7 of 12 opportunities, or 58.3%. Three disagreements concern whether a prompt began before the response window closed; two concern whether an AAC selection met the response definition. The team preserves both records, clarifies the rules, checks AAC access, and schedules a new independent sample.

Audit the denominator for Imani

The denominator for Imani is the eligible cohort defined before review. Report planned, due, observed, matched, invalid, missing, and open units separately. A percentage gains meaning from the unit, coverage, dates, conditions, and disagreement pattern.

Protect access and safety during the sample for Imani

The observation for Imani keeps AAC, interpreters, mobility, food, water, bathroom use, prescribed care, health supports, rest, relationships, and emergency help available. Observer procedure never delays immediate safety or mandated action. Record changes in access, health, assent, distress, visibility, staff behavior, or routine.

Use current sources within their scope for Imani

For Imani, the BACB ethics hub points to current ethics materials. The CASP public summary supplies high-level autism-treatment scope. The BACB Ethics Code addresses competence, client involvement, consent and assent when applicable, documentation, supervision, risk, and evaluation for covered behavior analysts. The BCBA Test Content Outline includes measurement, reliability, sampling, and data-based decisions as examination content.

For Imani and the structured low-IOA troubleshooting question, Vollmer, Sloman, and St. Peter Pipkin discuss practical implications of data reliability and treatment-integrity monitoring. Essig, Rotta, and Poling review IOA and fidelity reporting in an identified research corpus. Reed and Azulay describe agreement calculations and a calculation tool. ASHA says AAC users should always have access to their tools or devices. The responsible clinician still selects a case-specific measurement and review plan.

Close the IOA review for Imani

Review the low-IOA investigation log with Imani through accessible communication and with the responsible clinician. Record the question, sample, method, arithmetic, disagreement types, limits, repair, owner, next observation, and review date. Reopen the review when the definition, observer, setting, system, exposure, access, prevalence, procedure, or decision changes.

Before release, confirm that all five disagreements have a category, evidence, owner, and disposition. Check that access and system defects were repaired, revised rules have a version, and the next sample is independent. A low result remains a measurement-quality issue until evidence supports a more specific conclusion.

Related resources

Sources