Pragmatic Full Factorial Experiment — Real-World Multi-Factor Design
Pragmatic Full Factorial Experimental Design · Also known as: pragmatic factorial trial, real-world full factorial design, effectiveness full factorial experiment, pragmatic 2^k experiment
A pragmatic full factorial experiment combines the complete crossing of all factor levels (the full factorial structure) with the broad eligibility criteria, flexible delivery, and real-world conditions of a pragmatic trial. Every possible combination of factors is tested simultaneously, yielding both main effects and all interaction effects, while deliberately relaxing strict laboratory controls to reflect how interventions actually operate in practice.
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 pragmatic full factorial experiment when you need to evaluate two or more independent intervention components simultaneously in real-world settings and interaction effects between components are substantively important. It is especially valuable in behavioural, educational, public health, and health services research where interventions are multi-component and deployment context matters. It is NOT appropriate when sample size is too small to power all 2^k cells adequately; when factors are not genuinely independent or cannot be crossed ethically; when interactions are unlikely and a simpler design would suffice; or when tight experimental control is required (use a laboratory factorial design instead). With more than four or five binary factors the number of cells grows rapidly (32–64 cells) and a fractional factorial or sequential MOST/SMART design becomes more practical.
Strengths & limitations
- Simultaneously estimates all main effects and interaction effects from a single experiment — no need for separate studies per factor.
- Pragmatic framing ensures findings are directly relevant to real-world implementation and policy decisions.
- Efficient relative to running separate one-factor-at-a-time studies when factorial structure is appropriate.
- Interaction detection is a unique and policy-relevant strength — knowing that component A only works when component B is also present changes implementation strategy.
- Results transfer well to diverse populations and delivery contexts because eligibility is broad.
- Sample size requirements grow exponentially with the number of factors: 2^k cells each requiring adequate power is a serious practical constraint.
- Higher-order interactions (three-way and above) are often hard to interpret substantively and may be unstable with typical sample sizes.
- Pragmatic conditions increase outcome variance, which reduces power relative to a tightly controlled laboratory version of the same design.
- Maintaining treatment fidelity across many cells in real-world settings is operationally demanding.
- Results may be difficult to publish if journals expect the cleaner effect-size profiles typical of single-intervention RCTs.
Frequently asked
How is this different from a standard full factorial experiment?
A standard full factorial experiment typically applies tight laboratory-style controls — narrow eligibility, standardized delivery, close monitoring — to maximize internal validity. A pragmatic full factorial experiment uses the same complete crossing of all factors but deliberately adopts broad eligibility, flexible delivery, and real-world conditions to maximize external validity. The statistical model is the same; the design philosophy and resulting generalizability differ substantially.
How many participants do I need?
Each treatment cell must be adequately powered for the key comparisons. For a 2^3 (8-cell) design with a medium effect size and 80% power, you typically need at least 40–50 participants per cell, so 320–400 total. For higher-order interactions, power requirements are even larger. Use simulation or dedicated software (e.g., the MOST framework tools) to plan sample size, and consider a fractional factorial design if the full complement is not achievable.
What is the MOST framework and how does it relate?
The Multiphase Optimization Strategy (MOST), developed by Linda Collins and colleagues, is a principled framework for building and optimizing multi-component interventions. Its optimization phase frequently uses a full or fractional factorial experiment under pragmatic conditions to screen which components and interactions contribute to efficacy. MOST provides the broader decision-making context; the pragmatic full factorial is the specific design tool used in the optimization phase.
Can I use this design with cluster randomization?
Yes. When natural delivery units (schools, clinics, communities) cannot be individually randomized or when contamination across participants in the same unit is likely, clusters are randomized to factorial cells. The analysis must then use multilevel or mixed-effects models that account for clustering, and the number of clusters rather than individuals determines power for between-cluster comparisons.
When should I prefer a fractional factorial design instead?
When you have four or more factors and the sample size or budget makes a full factorial infeasible, a fractional factorial design tests a carefully chosen subset of all combinations. It can estimate all main effects and selected low-order interactions with far fewer cells. The cost is that some higher-order interactions are confounded (aliased) with lower-order ones. If your primary aim is screening rather than confirming all interactions, a fractional design is more practical.
Sources
- Thorpe, K. E., Zwarenstein, M., Oxman, A. D., Treweek, S., Furberg, C. D., Altman, D. G., ... & Chalmers, I. (2009). A pragmatic-explanatory continuum indicator summary (PRECIS): a tool to help trial designers. Journal of Clinical Epidemiology, 62(5), 464-475. DOI: 10.1016/j.jclinepi.2008.12.011 ↗
- Montgomery, D. C. (2017). Design and Analysis of Experiments (9th ed.). Wiley. ISBN: 978-1119492443
How to cite this page
ScholarGate. (2026, June 3). Pragmatic Full Factorial Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/pragmatic-full-factorial-experiment
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.
- Cluster Randomized Factorial ExperimentExperimental design↔ compare
- Factorial Randomized Controlled TrialExperimental design↔ compare
- Fractional Factorial ExperimentExperimental design↔ compare
- Full Factorial ExperimentExperimental design↔ compare
- Pragmatic Factorial ExperimentExperimental design↔ compare
- Pragmatic Randomized Controlled TrialExperimental design↔ compare