Pragmatic AB Design — Pragmatic AB Single-Case Experimental Design
Pragmatic AB Single-Case Experimental Design · Also known as: pragmatic single-case AB design, real-world AB design, AB phase design, naturalistic AB design
The Pragmatic AB Design is a single-case experimental design that collects repeated measurements of one individual or unit across two consecutive phases: a baseline phase (A) with no intervention, followed by an intervention phase (B). Deployed in real-world, clinically feasible conditions rather than tightly controlled laboratory settings, it is widely used in behavioral health, rehabilitation, education, and applied psychology to generate actionable evidence about individual-level treatment effects.
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 the Pragmatic AB Design when the research or clinical question concerns whether a specific intervention changes an outcome for an individual (or a small set of individuals) under real-world conditions, and an extended withdrawal of treatment (as required by ABAB designs) is ethically or practically unfeasible. It is appropriate in rehabilitation, behavioral health, education, and applied psychology settings where daily or session-level repeated measurement is feasible. Do NOT use it when the research goal is causal inference at the group level — AB designs lack the experimental controls (randomization, withdrawal, replication) to rule out maturation or history threats, making causal claims weak. If internal validity is the priority and withdrawal is permissible, prefer an ABAB or multiple-baseline design.
Strengths & limitations
- Feasible in routine clinical or educational practice without sacrificing participant welfare by withholding treatment indefinitely.
- Provides individual-level evidence that group-average findings cannot supply, making it directly relevant to practitioner decision-making.
- Repeated measurement yields a rich time-series picture of change trajectories that pre-post designs miss.
- Results can be aggregated across individuals using single-case meta-analytic methods (e.g., Tau-U, multilevel models) to build cumulative evidence.
- Low burden on participants and practitioners compared to randomized controlled trials in real-world settings.
- Without a withdrawal or replication phase, the AB design cannot rule out alternative explanations (maturation, history effects, regression to the mean), limiting causal inference.
- Findings are specific to the individual studied; statistical generalization to a population requires systematic replication across many participants.
- Visual analysis of time-series data is subject to analyst bias; low-variability baselines are required for reliable interpretation.
- The pragmatic (naturalistic) delivery of the intervention reduces control over confounding variables, weakening internal validity further.
Frequently asked
How is the Pragmatic AB Design different from a simple pre-post comparison?
A pre-post design takes one measurement before and one after intervention. The AB design takes many repeated measurements throughout both phases, producing a time-series that reveals level, trend, and variability. This richer data picture allows detection of gradual change trajectories and reduces the influence of any single outlier measurement.
Can I generalize findings from a Pragmatic AB Design to other people?
Not directly. The AB design produces findings specific to the individual studied. Generalization requires systematic replication — running the same study with multiple individuals and observing consistent effects. When replications accumulate, meta-analytic methods such as Tau-U can synthesize effect sizes across participants.
What effect size should I report for an AB design?
The most commonly recommended non-overlap statistics are PND (Percentage of Non-Overlapping Data), NAP (Non-Overlap of All Pairs), and Tau-U (which adjusts for baseline trend). Tau-U is currently the most methodologically defensible choice because it accounts for autocorrelation and baseline trends, and it is supported by available software (e.g., the Tau-U calculator by Parker et al.).
When should I upgrade from an AB design to an ABAB or multiple-baseline design?
Upgrade when causal inference is the goal and the intervention effect is reversible upon withdrawal (ABAB), or when you have multiple participants, settings, or behaviors that allow staggered introduction of the intervention (multiple-baseline). Use the AB design only when withdrawal is unethical or impractical and you accept the reduced causal certainty.
How many baseline data points are enough?
The general guideline is a minimum of five data points showing a stable level (flat trend, low variability) before introducing the intervention. If the baseline is still trending after five points, collect more data until stability is achieved. An increasing baseline trend in the expected direction of improvement makes interpreting any B-phase change virtually impossible.
Sources
- Kazdin, A. E. (2011). Single-Case Research Designs: Methods for Clinical and Applied Settings (2nd ed.). Oxford University Press. ISBN: 978-0195341881
- Tate, R. L., Perdices, M., Rosenkoetter, U., McDonald, S., Togher, L., Shadish, W., Horner, R., Kratochwill, T., Barlow, D. H., Kazdin, A., Sampson, M., Shamseer, L., & Vohra, S. (2016). The Single-Case Reporting Guideline In BEhavioural Interventions (SCRIBE) 2016 Statement. Physical Therapy, 96(7), e1–e10. DOI: 10.2522/ptj.2016.96.7.e1 ↗
How to cite this page
ScholarGate. (2026, June 3). Pragmatic AB Single-Case Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/pragmatic-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
- Single-Case Experimental DesignDisability Studies↔ compare