2.
Two Main Types of Variation
There are two main types of variation in a process:
A. Common Cause Variation (Natural Variation)
Definition: Variation that is inherent in the process. It is consistent and predictable
within statistical limits.
Characteristics:
o Random and small fluctuations.
o Present all the time in the process.
o No single cause can be identified.
o Process is considered “in control” if only common cause variation exists.
Example:
o Length of screws produced by a machine: ±0.1 mm around the target.
o Daily sales of a retail store fluctuate slightly due to normal customer behavior.
Implication: To improve the process, the system itself must be changed. Individuals
cannot be blamed.
B. Special Cause Variation (Assignable Variation)
Definition: Variation that is caused by specific, identifiable factors that are not part of
the process all the time.
Characteristics:
o Unusual or unexpected variation.
o Can be traced to a specific cause.
o Indicates the process may be “out of control.”
Example:
o A machine suddenly breaks and produces defective screws.
o A sudden spike in sales due to a marketing promotion or a one-time event.
o Human error, material defect, or environmental change.
Implication: Action can be taken to remove or control the cause.
1. What is Variation?
Variation is any difference in process output — it’s inevitable. Even with highly controlled
processes, perfect uniformity is impossible.
Variation can come from multiple sources and is critical in determining process capability
and product quality.
2. Types of Variation
Variation is generally categorized in two major ways:
A. By Cause
1. Common Cause Variation (Natural Variation)
o Inherent in the process.
o Small, random fluctuations.
o Predictable within control limits.
o Example:
A pen manufacturer: pen ink flow varies slightly, but all pens are
usable.
Temperature fluctuations in a factory: 24–25°C in an air-conditioned
plant.
2. Special Cause Variation (Assignable Variation)
o Arises from specific, identifiable causes.
o Large or unusual fluctuations.
o Signals that something is wrong in the process.
o Example:
Sudden power outage stops machines, creating defective products.
Raw material batch is faulty, causing product failure.
B. By Pattern (Observed in Data)
Variation can also be described visually:
1. Random Variation
o No predictable pattern.
o Usually common cause.
o Example: Slight differences in cookie weights (50.0, 50.2, 49.9 g).
2. Trend/Drift
o Gradual increase or decrease over time.
o Often indicates wear and tear or tool degradation.
o Example: A machine’s torque output slowly decreases over months.
3. Cycle
o Repeating pattern over time.
o Often related to environmental or operational cycles.
o Example: Higher electricity usage on weekdays than weekends.
4. Outliers
o Extreme values outside the normal range.
o Usually special cause variation.
o Example: One cookie weighing 55 g while others are ~50 g.
3. Sources of Variation
Sources can be categorized into 6M or 8M framework (common in quality engineering):
Source Example
Man Operator skill, fatigue
Machine Tool wear, calibration errors
Material Poor raw materials, batch inconsistencies
Method Different procedures, wrong setup
Measurement Gage error, imprecise measurement
Mother Nature (Environment) Temperature, humidity
Management Policy or scheduling changes (sometimes included)
4. Measuring Variation
Range (R): Difference between the largest and smallest values in a dataset.
Standard Deviation (σ): Average spread of data around the mean.
Variance (σ²): Square of the standard deviation, gives a measure of dispersion.
Control Charts: Visualize variation over time to identify common vs. special causes.
5. Why Variation Matters
Affects product quality and customer satisfaction.
Helps identify improvement opportunities.
Drives process control decisions (when to investigate, when to improve the system).