Intervention Selection From Assessment, Evidence, Preferences, and Contextual Fit

The Four Pillars of Defensible Selection

The BACB and Cooper, Heron, and Heward frame evidence-based practice (EBP) as the integration of (1) the best available evidence, (2) clinical expertise/judgment, and (3) client values and context. Domain H operationalizes this into four selection inputs you must weigh together:

1. Assessment results — what the FBA, skills assessment, preference assessment, records review, and direct observation reveal about function, repertoire, and barriers.
2. Scientific evidence — procedures with empirical support for similar behaviors, functions, and populations.
3. Client and stakeholder preferences — the values, goals, and acceptability judgments of the client, family, and team.
4. Contextual fit — whether real people, in the real setting, with real resources, can implement and sustain the plan with integrity.

The single most common exam error is treating these as a menu to pick from. They are filters applied together. An intervention that passes the evidence filter but fails the contextual-fit filter is not a good selection — it is a plan that will not be run as designed.

Start With Function, Not Topography

Intervention selection begins with the assessment, and for behavior reduction that means the function identified by functional assessment or functional analysis. Two children may both hit, but if one hits to escape demands and the other to gain attention, the correct interventions diverge completely. Function-based interventions — DRA of an escape mand, demand fading, and noncontingent escape for the escape-maintained child — are more efficient and more durable than topographically chosen procedures.

The distractors to reject are the procedures chosen for the wrong reason: because they are familiar to the team, fast to start, popular, or merely match the behavior's appearance. A common trap option recommends a default reductive procedure (e.g., time-out) without reference to function; for an escape-maintained behavior, time-out from positive reinforcement can actually reinforce escape and is wrong on its face.

Contextual Fit, Preferences, and the Least-Restrictive Lens

Contextual fit is where strong candidates separate from weak ones. A plan must match the people and place, not an ideal lab. Run the plan through these questions:

• Can implementers deliver the procedure at the required moments, every time, across the day?
• Are materials, staffing ratios, and schedules realistic in this setting?
• Does the plan respect the family's language, culture, dignity, and the client's right to choose?
• Can valid data be collected without disrupting instruction or care?
• Is the procedure the least restrictive option that is still likely to be effective, with risks and side effects monitored?

Client and stakeholder preferences are not a courtesy — they predict adherence and social validity. Procedures families find aversive or stigmatizing get dropped at home even when the clinic data look good. Where evidence is comparable, preference and acceptability should tip the choice.

Worked example. Two interventions for escape-maintained aggression have similar support: (A) a demanding errorless-teaching protocol requiring 1:1 staffing and continuous data, and (B) functional communication training (FCT) for a break plus brief task tolerance, runnable by the existing classroom aide. The classroom has one aide for six students and the family prioritizes Devon staying in the general-education room. Plan A is technically excellent but cannot be run with integrity here; Plan B fits the staffing, the setting, and the family's inclusion priority. The defensible selection is B — and a plan to build capacity if A's components are later needed.

Resolving Trade-Offs Between Evidence and Feasibility

The exam rarely offers a free choice where one option is best on every dimension. More often you must trade evidence strength against feasibility, risk, or values. Two decision rules cover most items:

1. When evidence is comparable, choose better fit, lower risk, and stronger buy-in. Equivalent procedures are differentiated by whether they will actually be implemented.
2. When the best-supported procedure cannot be implemented well, adapt it or build capacity first — do not deploy it on faith. A perfectly evidenced plan run at 40% integrity produces worse outcomes than a slightly less optimal plan run at 95% integrity.

What you should never do is select on convenience alone ("we already do time-out for everyone") or on evidence alone while ignoring that the setting cannot support it. The credited Domain H answer integrates the pillars: it cites the assessment-derived function, references the evidence base, respects client values, and confirms the setting can run it. That integration — not any single input — is the hallmark of a defensible selection.

There is also a hierarchy among the pillars when they conflict on a safety dimension. A procedure that is well-evidenced but poses an unacceptable risk in this context (for example, a physically managed procedure with untrained staff) is not selected just because the literature supports it; risk and least-restrictiveness gate the choice. Conversely, a preferred, low-risk procedure that the assessment shows is contraindicated by function (reinforcing the wrong contingency) is also rejected, however acceptable the family finds it.

In practice, sequence the filters: first confirm the procedure addresses the assessed function and is acceptably safe/least-restrictive; then confirm it has evidence for this function and population; then confirm client values and contextual fit. A procedure must clear all four. When exam options each emphasize one pillar, the correct choice is the one that survives every filter — not the option that maximizes a single dimension while quietly failing another.