Overview
ISO/TR 12845:2010 - Selected illustrations of fractional factorial screening experiments is an informative Technical Report from ISO that describes the practical steps to design, run and analyse two‑level fractional factorial designs. Rather than prescribing new rules, it illustrates methodology through six detailed, real‑world examples, showing how fractional factorial screening experiments can identify important factors efficiently when full factorial designs are impractical.
Key topics
- Step‑wise experimental workflow: the report presents a seven‑step framework (define objectives; choose response variable(s); identify factors; select an appropriate fractional factorial design; analyse results with numerical summaries and graphical displays; present findings; run confirmation experiments).
- Two‑level fractional factorial designs: practical guidance on selecting and using fractional designs for screening important main effects and low‑order interactions.
- Design resolution and aliasing: explanation of resolution (III, IV, V) and how aliasing affects interpretation of main effects and interactions.
- Analysis methods: use of ANOVA, graphical diagnostics and software‑based analysis to summarise and interpret experimental outcomes.
- Experiment structure and terminology: definitions (factors, levels, blocks, replication, randomization, split‑plot), symbols and abbreviated terms consistent with ISO/3534 series.
- Practical nuances: discussion of issues such as blocking, centre points, replication, and confirmation runs that commonly arise in applied experiments.
- Worked examples: six annexes (A–F) offering applications across diverse domains to illustrate methodological choices and interpretation.
Applications
ISO/TR 12845:2010 is focused on screening experiments where resource constraints make full factorial testing impractical. Typical applications include:
- Rapid factor screening in product development and process optimization
- Early‑stage Six Sigma improvement projects (Black Belts / Green Belts)
- Process validation and formulation optimization in pharmaceuticals and polymers
- Experimental designs in marketing, manufacturing, materials science and biological studies
Representative example topics from the report include direct mail marketing, polymer emulsion optimization, PVC foam formulations, insulin process validation, washing machine testing and a bacterium aggregation study.
Who should use it
- Statisticians and experiment designers applying Design of Experiments (DoE)
- Quality engineers, process engineers and R&D scientists conducting screening studies
- Six Sigma practitioners seeking rigorous, software‑oriented examples
- Regulators and auditors wanting an authoritative reference for experimental practice
Related standards
ISO/TR 12845:2010 references ISO 3534 parts 1–3 (statistics vocabulary and design of experiments terminology) and aligns with international statistical terminology. Note: as a Technical Report, ISO/TR 12845 is informative and illustrative rather than a prescriptive normative standard.
Keywords: ISO/TR 12845:2010, fractional factorial designs, two‑level designs, screening experiments, design of experiments, DoE, ANOVA, design resolution, Six Sigma.