Hybrid Full Factorial Design — Combined Factorial Experimental Strategy
Hybrid Full Factorial Experimental Design · Also known as: hybrid factorial design, mixed full factorial design, combined factorial design, HFFD
Hybrid full factorial design is an experimental strategy that applies a full factorial structure to a selected subset of factors — those believed to have the strongest interactions — while treating remaining factors with a reduced or fractional scheme. This hybrid approach balances the complete interaction information of a full factorial with the run-count efficiency of fractional designs, making it practical for studies with many factors where a pure full factorial would be prohibitively expensive.
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 hybrid full factorial design when you have five or more factors and a moderate-to-large experimental budget, some factors are known or strongly suspected to interact, and a pure full factorial is unaffordable. It is well suited to engineering optimization, product formulation, and process development where a few critical factors dominate and the rest can be studied with lower resolution. Do not use it when you have no prior knowledge to guide the factor partition — if all factors are equally suspect, a fractional factorial or definitive screening design is more appropriate. Also avoid it when the response surface is highly nonlinear across all factors simultaneously, as the hybrid structure may miss curvature terms in the fractional portion.
Strengths & limitations
- Captures all main effects and interactions among the most important factors without sacrificing every combination of secondary factors.
- Substantially reduces the total number of runs compared to a full factorial across all factors, improving feasibility.
- Flexible: the core-factor subset and the fractional scheme for secondary factors can be tailored to the available budget.
- Supports hierarchical analysis — detailed inference on core factors and screening-level inference on secondary factors within one study.
- Compatible with standard ANOVA and regression analysis software.
- Requires prior knowledge or subject-matter judgment to correctly partition factors into core and secondary groups; a wrong partition may leave important interactions unresolved.
- The fractional portion for secondary factors still carries aliasing, so interaction effects among secondary factors may be confounded.
- Design construction is more complex than a standard factorial or fractional factorial and typically requires specialist statistical software.
- If the wrong factors are assigned to the fractional portion, confirmation runs are needed to resolve ambiguity, potentially erasing the cost savings.
Frequently asked
How is a hybrid full factorial design different from a split-plot design?
A split-plot design addresses practical constraints on randomization — some factors are hard to change and are held constant across groups of runs (whole plots), while others change freely (subplots). A hybrid full factorial design, by contrast, is about resolution: some factors receive full factorial coverage and others receive fractional coverage, but randomization is applied uniformly across all runs. The motivation is cost and run-count efficiency, not physical constraints on factor changing.
How do I decide which factors go into the full factorial portion?
Use prior knowledge, mechanistic understanding, or results from a preliminary screening study (e.g., Plackett-Burman or definitive screening design). Factors with known or suspected two-way or higher-order interactions, or factors whose effect on the response is critical to the study objective, should be assigned to the full factorial core. When in doubt, err on the side of including a factor in the core group.
What software supports hybrid full factorial designs?
JMP (Custom Design platform), Minitab (Design of Experiments module), and R packages such as FrF2 and DoE.base support construction of combined and hybrid designs. The analyst typically builds the two sub-matrices separately and then merges them, or uses a custom-design algorithm that allows specifying different resolution requirements per factor group.
Can I add center points to a hybrid full factorial design?
Yes. Center points are most meaningful in the full factorial portion where the factor space is fully covered, and they allow estimation of curvature among the core factors. Adding center points to the fractional block is also possible but provides limited curvature information given the sparser coverage of that region.
Is the hybrid full factorial design suitable for screening studies?
Not as a primary screening tool. Pure screening designs (Plackett-Burman, definitive screening) are more efficient when you have many factors and no a priori grouping. The hybrid design is best used when screening has already been done and a subset of important factors has been identified for deeper investigation alongside some additional variables.
Sources
- Montgomery, D. C. (2017). Design and Analysis of Experiments (9th ed.). Wiley. ISBN: 978-1119113478
- Antony, J. (2014). Design of Experiments for Engineers and Scientists (2nd ed.). Elsevier. ISBN: 978-0080994178
How to cite this page
ScholarGate. (2026, June 3). Hybrid Full Factorial Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/hybrid-full-factorial-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.
- Central Composite DesignExperimental design↔ compare
- Latin Hypercube SamplingSimulation↔ compare
- Response Surface MethodologyExperimental design↔ compare