Multi-Vari Studies separate variation within pieces, between pieces, and over time so teams can locate dominant sources of process variation.
Definition
A Multi-Vari Study is a structured variation study that helps identify whether process variation occurs within units, between units, or over time. It uses planned sampling and graphical analysis to locate the dominant family of variation before deeper analysis.
It is useful when a process output varies but the team does not yet know whether the cause is positional, part-to-part, time-related, machine-related, or lot-related.
History
Multi-vari methods grew from industrial statistics and Six Sigma practice as a practical way to narrow sources of variation. They are often used before DOE or detailed root cause testing.
When to Use
Use multi-vari studies when defects or measurements vary by location, cavity, station, part, time, batch, shift, tool, or setup. It is especially useful for dimensional, strength, temperature, coating, fill, or process-output variation.
Step-by-Step
- Define the response and suspected variation families.
- Plan sampling across within-unit, between-unit, and time dimensions.
- Validate measurement system adequacy.
- Collect data in sequence with context.
- Plot multi-vari charts or grouped comparisons.
- Identify dominant variation family.
- Focus root cause analysis on that family.
- Verify causes through data, observation, or experiments.
Examples
- Molding: Variation differs by cavity more than by time.
- Coating: Thickness varies across part location.
- Machining: Measurements drift over time after tool wear.
Common Pitfalls
- No clear sampling plan.
- Weak measurement system.
- Mixing too many products or conditions.
- Skipping graphical review.
- Jumping to DOE before locating variation family.
- Ignoring process sequence.