502. Basic Statistics I – Definitions of Commonly Used Statistical Terms

When working with quality management, Six Sigma methodologies, or process control, statistical language forms the backbone of problem-solving. Without a clear grasp of basic statistical terminology, interpreting process performance, variations, and capabilities becomes nearly impossible.

Many of these statistical concepts appear in daily conversation, but in quality engineering and data analysis, they carry precise mathematical definitions. This article serves as Part 1 of our comprehensive statistical series, establishing the core definitions required for statistical process control (SPC).

Measures of Central Tendency & Data Characteristics

Central Tendency: A numerical measure describing the central point around which a group of data values clusters. The most common measures are the mean, median, and mode.
- Average (Mean): The arithmetic average of a dataset. Calculated by summing all sample values (S $\sum x$ ) and dividing by the total number of elements ( $n$ ) in the sample $$\bar{x} = \frac{\sum x}{n}$$
- Median: The middle value in a dataset when all observations are arranged in sequential ascending or descending order.
- Mode: The specific data point or value that occurs with the highest frequency in a given dataset.
Characteristic: A measurable and monitorable physical or operational property of a process input or output (e.g., length, weight, temperature, or cycle time).
Cycle Time: The total elapsed time required from the moment a process, task, or service begins until it reaches final completion.

Measures of Dispersion & Variability

Understanding data spread is critical to reducing defects in Six Sigma projects.

Term	Definition	Practical Application
Variability	The inherent property of a process, characteristic, or system to yield different values upon repetition.	Identifies process instability.
Range ( $R$ )	The simple difference between the largest ( $X_{max}$ ) and smallest ( $X_{min}$ ) values in a dataset.	Quick measure of data spread in small samples.
Variance (σ² or s²)	The arithmetic average of squared deviations from the mean.	Mathematical foundation for advanced statistics.
Standard Deviation (σ or s)	The square root of the variance, representing data dispersion in original measurement units.	Primary metric for calculating Six Sigma capability.
Long-term Variation	The observed variation of a process characteristic over extended periods, capturing all environmental and operational shifts.	Evaluates true long-term process capability.
Variables & Variable Data	Continuous data that can be meaningfully subdivided into decimal increments (e.g., dimensions, pressure).	Used in continuous measurement charts ( $X$ -bar and $R$ ).

Control Limits vs. Specification Limits

A major Point of confusion in quality management is distinguishing between customer requirements and natural process boundaries:

Lower Specification Limit (LSL) & Upper Specification Limit (USL): The absolute boundaries set by customer requirements, engineering drawings, or statutory standards. Values outside these limits constitute non-conformities or defects.
Lower Control Limit (LCL) & Upper Control Limit (UCL): Statistically calculated boundaries (typically located $3\sigma$ below and above the process mean) representing the natural variation of a process in state of statistical control.
Trend: A gradual, systematic, and predictable pattern of change in process data plotted over time or across consecutive production runs.

Summary Comparison: Specifications vs. Control Limits

                   [ Customer Specification Limits (LSL / USL)]
                             (Voice of the Customer)
                                       │
            ┌──────────────────────────┴──────────────────────────┐
            ▼                                                     ▼
  Upper Control Limit (UCL) <─── Natural Process Spread ───> Lower Control Limit (LCL)
                             (Voice of the Process)

By standardizing these statistical terms within your team, you establish a common language essential for conducting root-cause analysis, calculating $C_p$ / $C_{pk}$ indices, and implementing statistical process control.

Written by Ravi Prakash—Quality Expert (38+ yrs exp). Connect on LinkedIn or Contact Us.

Support Our Website

Your support helps us continue delivering good content and useful data. If you’d like to contribute, here’s how:

From anywhere in the world (PayPal): https://www.paypal.com/paypalme/rpbehara
From India (UPI): speakingdata@ybl

The Next chapter is on Basic Statistics 2

502. Basic Statistics-I-Definitions of commonly used Statistical Terms