Pilot AB Design — Preliminary AB Single-Subject Experimental Design
Pilot AB Single-Subject Experimental Design · Also known as: pilot AB phase design, preliminary AB design, exploratory AB single-case design, feasibility AB design
A pilot AB design applies the two-phase baseline-then-intervention structure of the AB single-subject design in an explicitly exploratory or feasibility mode — before committing to a more rigorous reversal or multiple-baseline study. The researcher collects repeated baseline (A) and intervention (B) data from one or a few individuals primarily to test measurement procedures, estimate effect size, verify data stability, and determine whether a stronger single-case design is warranted and feasible.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use a pilot AB design when you are entering a new single-subject research program and need to test the feasibility of measurement procedures, verify baseline stability, and obtain a preliminary effect-size estimate before committing participants or resources to a more controlled design. It is appropriate when the behavior or clinical context is novel enough that standard parameters are unknown, when a reversal is anticipated to be ethically or practically problematic and you need to gather preliminary data before justifying a multiple-baseline design, or when training new observers or testing a new outcome measure. Do not use the pilot AB design as the sole or final evidence for causal claims about an intervention: without a withdrawal or replication phase, history and maturation threats remain uncontrolled. If the question is whether the intervention works — not whether studying it is feasible — use ABA, ABAB, or multiple-baseline designs from the outset.
Strengths & limitations
- Provides a low-burden, practically feasible entry point for testing the viability of a single-subject research program before full-scale implementation.
- Allows calibration of measurement tools, training of observers, and verification of baseline stability under real-world conditions.
- Yields preliminary effect-size and variability estimates that inform sample-size and phase-length decisions for a subsequent stronger design.
- Ethically acceptable as an initial phase when reversing or withholding a potentially beneficial intervention for a full reversal design is premature.
- Produces actionable go/modify/stop information that prevents wasted resources on a poorly designed full study.
- No experimental control: the AB structure cannot rule out history, maturation, or regression to the mean as explanations for observed change.
- Results from a single or small number of individuals have no statistical generalizability; replication across participants and settings is necessary before broader claims can be made.
- Pilot data may overestimate effect sizes due to novelty effects, observer reactivity, and the non-representative conditions of a pilot phase.
- As a pilot, the design is explicitly not powered or structured to confirm efficacy, yet findings are sometimes misrepresented as evidence that the intervention works.
Frequently asked
How is a pilot AB design different from a standard AB design?
The underlying phase structure is identical — baseline followed by intervention. The difference is the explicit intent and interpretive frame. A pilot AB design is conducted specifically to answer feasibility questions (Can we measure this reliably? Is the baseline stable? Is the logistical setup workable?) rather than to document the effect of the intervention per se. Reporting a pilot AB design involves evaluating those feasibility criteria and using the findings to plan a stronger follow-up design.
Can I publish findings from a pilot AB design?
Yes, but the pilot framing must be explicit and conclusions must be appropriately modest. Suitable publication formats include feasibility reports, preliminary investigations, or single-case pilot sections within a larger study. The paper should clearly state that causal inferences cannot be drawn, that findings are intended to guide a subsequent controlled study, and that the primary outcomes are feasibility metrics rather than intervention efficacy.
When should I move directly to ABA or ABAB instead of starting with a pilot AB?
If you already have a well-validated measurement tool, reliable baseline data from prior work, and a clear understanding of the logistical context, a pilot AB phase is unnecessary — proceed directly to the stronger design. Pilot AB phases are most valuable when you are entering genuinely new territory where measurement procedures, baseline variability, or delivery feasibility are uncertain.
How many participants should a pilot AB study include?
Typically one to three participants, sufficient to test the measurement and logistical setup across a small range of individual variation. The goal is not statistical power but operational confidence that the protocol works. If the pilot involves more than one participant, structured replication logic — staggering start times, for example — begins to approximate a multiple-baseline design and provides a stronger platform for subsequent work.
Is inter-observer reliability required in a pilot AB design?
Yes — and the pilot phase is an ideal time to establish and report it. Calculating inter-observer agreement (typically percentage agreement or kappa) during the pilot validates the measurement system before the full study begins. Conventional minimums are 80% or higher for percentage agreement; if reliability falls below this threshold in the pilot, revise the operational definition or observer training before proceeding.
Sources
- Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1(1), 91-97. DOI: 10.1901/jaba.1968.1-91 ↗
- Kazdin, A. E. (2011). Single-Case Research Designs: Methods for Clinical and Applied Settings (2nd ed.). Oxford University Press. ISBN: 978-0195341881
How to cite this page
ScholarGate. (2026, June 3). Pilot AB Single-Subject Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/pilot-ab-design
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- AB DesignExperimental design↔ compare
- ABA DesignExperimental design↔ compare
- ABAB designExperimental design↔ compare
- Multiple Baseline DesignExperimental design↔ compare
- Pilot Randomized Controlled TrialExperimental design↔ compare
- Single-Subject Experimental DesignExperimental design↔ compare