Opening the book…
With the independent values identified, the team can summarize its results.
Descriptive statistics describe the data actually observed. The mean uses every value and can be influenced strongly by an unusual observation. The median is the middle value after the values are ordered, or the average of the two middle values when their number is even. Examining both can help reveal an uneven pattern.
The sample standard deviation, or SD, describes how widely values are spread around the sample mean. It uses the same units as the original measurement. To calculate it, find each value’s difference from the mean, square each difference, add the squares, divide by n − 1, and take the square root. Here, n is the number of values. Dividing by n − 1 is part of estimating population variation from a sample.
Mean = sum of values ÷ n.
Sample SD = square root of [sum of squared deviations from the mean ÷ (n − 1)].
Worked example: summarizing scratch closure
Five independent culture preparations give vehicle-control closure values of 40%, 45%, 50%, 55%, and 60% at the specified time. Treat these as five independent preparation-level values for this example.
Mean = (40 + 45 + 50 + 55 + 60) ÷ 5 = 50% closure.
Deviations from the mean are −10, −5, 0, 5, and 10 percentage points. Their squared values sum to 100 + 25 + 0 + 25 + 100 = 250.
Sample SD = square root of (250 ÷ 4) = 7.91 percentage points.
Report the result as mean closure 50.0%, SD 7.9 percentage points, n = 5 independent preparations. The SD describes spread in closure values, not a 7.9% relative change in closure. Preserve the individual values so the reader can see the pattern.
|
Closure (%) |
Deviation (percentage points) |
Squared deviation |
|
40 |
−10 |
100 |
|
45 |
−5 |
25 |
|
50 |
0 |
0 |
|
55 |
5 |
25 |
|
60 |
10 |
100 |
|
Sum: 250 |
Sum: 0 |
Sum: 250 |
The coefficient of variation, or CV, expresses SD relative to the mean. For positive measurements on a scale with a meaningful zero, percent CV = (SD ÷ mean) × 100. For these closure values, CV = (7.91 ÷ 50) × 100 = 15.8%. It describes relative spread. Whether that spread is acceptable requires a criterion suited to the method and use.
CV can help compare variation at different positive concentration levels. It becomes unstable when the mean is near zero and is unsuitable for scales without a meaningful zero. For example, Celsius temperatures and qPCR Cq values should not be treated like concentrations in this calculation. Background-corrected measurements that cross zero also need a different approach.
An SD from two or three values is itself uncertain. More decimal places do not make that estimate more reliable. Use a sensible number of reported digits while retaining the original precision in the data record. For skewed data, a median and a clearly defined spread measure may describe the observations more usefully than a mean alone.
Attribution: Images AI-generated using ChatGPT Modified by Emalee Mackenzie, Irvine Valley College. Licensed under CC BY-NC-SA
This page titled 15.8: Mean Standard Deviation and Coefficient of Variation was last modified on Wed, 21 Oct 2026 01:35:51 GMT and is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Emalee MacKenzie.
The book could not be loaded just now.