Sensitivity Analysis-Integrated Design of Experiments
Also known as: SA-DoE, SA-integrated DoE, DoE with sensitivity screening, factor screening with sensitivity analysis
Sensitivity Analysis-Integrated Design of Experiments (SA-DoE) combines systematic experimental planning with formal sensitivity analysis to identify which input factors most strongly influence a response, then efficiently characterises those factors' effects. By embedding sensitivity screening into the DoE workflow, experimenters avoid wasting trials on inert variables and focus resources on the factors that truly drive system behaviour — making it especially valuable in simulation studies, product engineering, and complex process optimisation.
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
SA-DoE is the right choice when the number of candidate factors is large (typically more than five) and running a full factorial is cost-prohibitive, when experiments involve expensive physical tests or computationally intensive simulations, or when prior knowledge about factor importance is limited. It is particularly well-suited to engineering design, process optimisation, environmental modelling, and simulation-based studies. Avoid it when the factor set is already small and well-understood (a straightforward DoE suffices), when sensitivity analysis assumptions (e.g., factor independence) are strongly violated, or when the response is non-smooth in ways that invalidate variance-based decomposition.
Strengths & limitations
- Reduces experimental effort by eliminating inert factors before the main DoE, lowering cost and time.
- Provides a principled, quantitative basis for factor prioritisation rather than relying on domain intuition alone.
- Compatible with both physical experiments and computer simulation experiments (e.g., finite element, CFD models).
- Improves model quality by focusing the fitted response surface on the factors with genuine explanatory power.
- The sensitivity indices produced during screening are themselves informative outputs, not just a filtering tool.
- Requires an initial investment in sensitivity analysis runs before the main experiment begins, adding upfront cost.
- Sobol variance-based indices assume factor independence; correlated inputs require more advanced estimators.
- If the screening phase uses too few samples, important factors may be misclassified as inert and incorrectly dropped.
- The integrated workflow demands expertise in both DoE and sensitivity analysis, which are often siloed disciplines.
Frequently asked
What sensitivity analysis method should I use for the screening phase?
For a quick, model-free screening of many factors, Morris OAT (elementary effects) is computationally cheap and robust. For a more rigorous variance decomposition that separates main effects from interactions, Sobol indices are preferred but require more runs. In practice, Morris screening is common as a first-pass filter, with Sobol analysis reserved for the refined factor set or the fitted surrogate.
How many runs are needed for the sensitivity screening phase?
Morris screening needs r × (k + 1) runs, where r is the number of trajectories (typically 10–20) and k is the number of factors. Sobol-based methods need N × (k + 2) runs, where N is a base sample size (often 500–2000 for reasonable convergence). The screening investment is justified when it allows a substantially smaller main DoE.
Can I use SA-DoE with physical experiments rather than simulations?
Yes, but the upfront screening cost is higher for physical runs than for simulations. In practice, teams often use a low-fidelity computational model for screening and reserve physical experiments for the main DoE phase on the reduced factor set. If physical screening runs are feasible, a Plackett-Burman or Resolution III fractional factorial can serve a similar filtering role at lower run counts than a full SA.
What if the sensitivity screening and the DoE give contradictory importance rankings?
Contradictions usually signal that the initial factor ranges were mis-specified, that the screening sample was too small to converge, or that the response is strongly nonlinear in ways the screening design did not capture. Revisit the factor ranges, increase the screening sample, and consider a non-parametric sensitivity estimator before concluding that a factor is truly inert.
Is SA-DoE the same as a Plackett-Burman screening design?
No. Plackett-Burman and similar resolution III designs are DoE-only screening tools that estimate main effects by aliasing interactions. SA-DoE is a two-phase framework: the first phase uses sensitivity indices (which are not based on a linear model) to rank factors, and the second phase applies a chosen DoE to the retained factors. SA-DoE is more flexible and can handle nonlinear responses that would confound Plackett-Burman estimates.
Sources
- Saltelli, A., Tarantola, S., Campolongo, F., & Ratto, M. (2004). Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models. Wiley. ISBN: 9780470870938
- Montgomery, D. C. (2017). Design and Analysis of Experiments (9th ed.). Wiley. ISBN: 9781119113478
How to cite this page
ScholarGate. (2026, June 3). Sensitivity Analysis-Integrated Design of Experiments. ScholarGate. https://scholargate.app/en/experimental-design/sensitivity-analysis-integrated-design-of-experiments
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.
- Latin Hypercube SamplingSimulation↔ compare
- Response Surface MethodologyExperimental design↔ compare
- Taguchi MethodExperimental design↔ compare