Pilot Fractional Factorial Experiment
Also known as: pilot FFE, screening pilot design, pilot fractional factorial, pilot FF screening study
A pilot fractional factorial experiment is a small-scale preliminary study that uses a fractional factorial design — testing only a subset of all possible factor combinations — to screen multiple factors simultaneously before committing to a full-scale investigation. It provides early estimates of effect sizes, variance, and feasibility at substantially reduced cost and participant burden compared to a full factorial pilot or a full-scale trial.
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 fractional factorial experiment when you need to screen several factors (typically 4–8) simultaneously in a resource-limited preliminary phase, particularly when the direction and magnitude of effects are unknown and a full factorial pilot would be prohibitively expensive. It is especially appropriate in industrial process optimization, early-phase behavioral intervention development, and multi-factor clinical feasibility studies. Do not use it as a standalone confirmatory tool — the confounding structure of fractional designs means that individual effect estimates cannot be interpreted in isolation without further experimentation. It is also inappropriate when the number of factors is very small (2–3) and a full factorial is feasible, or when the primary goal is to establish efficacy with controlled error rates rather than screen factors.
Strengths & limitations
- Screens many factors simultaneously with far fewer experimental runs than a full factorial design.
- Generates feasibility data (recruitment rates, protocol adherence, variance estimates) essential for powering a full-scale study.
- Identifies which factors are worth including in a confirmatory study, avoiding wasted resources on inactive predictors.
- Structured design matrix ensures systematic coverage of the factor space rather than ad hoc exploration.
- Effect estimates from the pilot can directly inform sample size calculations for the main trial.
- The confounding (aliasing) structure of fractional designs means that some main effects and interactions cannot be separated without additional runs.
- Small cell sizes in pilot phases yield imprecise effect estimates with wide confidence intervals — sufficient for screening but not for inference.
- Does not provide the statistical power needed for confirmatory hypothesis testing; results must be treated as preliminary.
- Feasibility findings from a pilot may not generalize if the pilot population or setting differs from the intended full-scale context.
Frequently asked
How many runs do I need for a pilot fractional factorial experiment?
The minimum number of runs is determined by the chosen fraction and number of factors. A 2^(k-p) design requires 2^(k-p) runs. For example, screening 5 factors with a 2^(5-2) design requires only 8 runs. In practice, pilot studies often replicate some or all runs to get a variance estimate, but even 8–16 runs can screen 5–8 factors meaningfully.
What is resolution and why does it matter in a pilot?
Resolution describes which effects are confounded (aliased) with each other. Resolution III designs confound main effects with two-factor interactions; Resolution IV confounds two-factor interactions with each other but not with main effects. For a pilot focused on screening, Resolution III is often acceptable. For a pilot where two-factor interactions are plausible and important, Resolution IV is preferable.
Can I use the pilot data to power the full study?
Yes — and this is one of the primary purposes of a pilot. The variance estimate and preliminary effect size from the pilot feed into a formal power calculation for the confirmatory study. However, use this estimate cautiously: pilot variance estimates are imprecise, so it is advisable to be conservative or use a sensitivity analysis over a range of plausible variances.
How is this different from a full pilot randomized controlled trial?
A pilot RCT typically tests a single treatment versus control. A pilot fractional factorial experiment simultaneously tests multiple factors, allowing you to screen which factors are worth including in a confirmatory RCT. The two designs serve complementary purposes: use the fractional factorial to identify the active factors, then design a confirmatory RCT (or full factorial) around those factors.
Is it appropriate to report p-values from a pilot fractional factorial study?
Reporting p-values is technically possible but commonly discouraged for pilot studies, as the study is not powered for hypothesis testing. Effect sizes and confidence intervals with explicit caveats about precision are more informative and less likely to be misinterpreted as confirmatory evidence.
Sources
- Montgomery, D. C. (2017). Design and Analysis of Experiments (9th ed.). Wiley. ISBN: 978-1119492443
- Thabane, L., Ma, J., Chu, R., Cheng, J., Ismaila, A., Rios, L. P., Robson, R., Thabane, M., Giangregorio, L., & Goldsmith, C. H. (2010). A tutorial on pilot studies: The what, why and how. BMC Medical Research Methodology, 10(1), 1. DOI: 10.1186/1471-2288-10-1 ↗
How to cite this page
ScholarGate. (2026, June 3). Pilot Fractional Factorial Experiment. ScholarGate. https://scholargate.app/en/experimental-design/pilot-fractional-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.
- Factorial ExperimentExperimental design↔ compare
- Fractional Factorial ExperimentExperimental design↔ compare
- Full Factorial ExperimentExperimental design↔ compare
- Pilot Randomized Controlled TrialExperimental design↔ compare
- Response Surface MethodologyExperimental design↔ compare