What Is Descriptive Statistics and How Do You Use It in Quality Control?

If you have ever calculated an average, looked at the spread of your measurements, or checked whether your process is centered, you have already used descriptive statistics. In quality control, descriptive statistics is the first step to understanding your data before any advanced analysis. This article explains what it is, how to compute the key measures, and how to apply them correctly.

What It Is

Descriptive statistics is the branch of statistics that summarizes and describes the main features of a dataset. Unlike inferential statistics, which draws conclusions about a larger population from a sample, descriptive statistics simply reports what the data show. It gives you a compact picture of location, spread, and shape.

The authoritative definitions of these statistical terms are provided in ISO 3534-1:2006, Statistics — Vocabulary and symbols. For quality engineering practice, the conventions follow standard textbooks such as Montgomery's Introduction to Statistical Quality Control (Wiley). Descriptive statistics is the foundation for control charts, capability analysis, and hypothesis testing.

How It Works: Key Measures and Formulas

Descriptive statistics is built on three groups of measures:

### 1. Measures of Central Tendency (Location)

• Mean (arithmetic average): For a sample of \(n\) observations \(x_1, x_2, \dots, x_n\):

\[
\bar{x} = \frac{1}{n}\sum_{i=1}^{n} x_i
\]

• Median: The middle value when data are sorted in ascending order. If \(n\) is even, it is the average of the two middle values.
• Mode: The value that occurs most frequently.

### 2. Measures of Dispersion (Spread)

• Range: \(R = x_{\max} - x_{\min}\)
• Sample variance:

\[
s^2 = \frac{1}{n-1}\sum_{i=1}^{n}(x_i - \bar{x})^2
\]
Note the denominator is \(n-1\), not \(n\). This is the unbiased estimator for the population variance.

• Sample standard deviation: \(s = \sqrt{s^2}\). This is the most commonly reported measure of spread in quality work.

### 3. Measures of Shape

• Skewness: Indicates asymmetry. Positive skew means a long right tail; negative skew means a long left tail.
• Kurtosis: Indicates tailedness. High kurtosis means more extreme outliers than a normal distribution.

For a normal distribution, skewness is 0 and excess kurtosis is 0.

A Worked Illustrative Example

Example data (illustrative only): Suppose you measure the diameter (in mm) of 10 parts from a production line:

\[
10.2, \; 10.4, \; 10.1, \; 10.3, \; 10.2, \; 10.5, \; 10.3, \; 10.2, \; 10.4, \; 10.6
\]

Step 1 — Mean:
\[
\bar{x} = \frac{10.2+10.4+10.1+10.3+10.2+10.5+10.3+10.2+10.4+10.6}{10} = \frac{103.2}{10} = 10.32
\]

Step 2 — Median: Sorted data: 10.1, 10.2, 10.2, 10.2, 10.3, 10.3, 10.4, 10.4, 10.5, 10.6. The two middle values are 10.3 and 10.3, so median = 10.3.

Step 3 — Mode: 10.2 appears three times, so mode = 10.2.

Step 4 — Range: \(10.6 - 10.1 = 0.5\) mm.

Step 5 — Sample variance and standard deviation:
First compute squared deviations from the mean. The sum of squared deviations is 0.256. Then:
\[
s^2 = \frac{0.256}{10-1} = 0.0284
\]
\[
s = \sqrt{0.0284} \approx 0.169 \text{ mm}
\]

Interpretation (illustrative): The process center is about 10.32 mm, with typical variation of ±0.17 mm. The mean (10.32) is slightly above the median (10.3), suggesting a very mild right skew — but with only 10 points, this is not conclusive.

Remember: these numbers are for demonstration only. Do not treat them as real process conclusions.

Common Pitfalls

• Using \(n\) instead of \(n-1\) in the sample variance. This underestimates the true variability.
• Reporting the mean without the standard deviation. Location alone is misleading.
• Ignoring outliers before summarizing. A single extreme value can distort the mean and standard deviation.
• Over-interpreting skewness and kurtosis from small samples. With fewer than 30 points, these measures are unstable.
• Confusing sample statistics with population parameters. Descriptive statistics describes your sample; it does not prove what the whole population looks like.

Closing

Descriptive statistics turns raw numbers into actionable insight — it tells you where your process is centered, how much it varies, and whether the distribution looks normal. For a quick and accurate calculation of mean, median, mode, range, variance, standard deviation, skewness, and kurtosis, use the free tool at 6SQ Descriptive Statistics Calculator. Enter your data, get the summary in seconds, and start your quality analysis on solid ground.