A. Advanced Central Tendency
While Mean, Median, and Mode are standard, biological research requires knowing exactly when to use specific variations.
Critical Property:
The sum of deviations of a set of observations from their arithmetic mean is always zero: Σ (Xᵢ - X̄) = 0.
When to use: Ideal for symmetric, normally distributed data without extreme outliers (e.g., standard physiological parameters like blood glucose levels in a controlled lab strain).
When to use: Mandatory for exponential growth or ratio data. If you are tracking the growth rate of a bacterial culture, population genetics frequencies, or tissue doubling times, the arithmetic mean will give a biased, overinflated result. You must use the Geometric Mean.
When to use: Used when dealing with rates, ratios, or time-per-unit metrics. For example, calculating the average speed of an enzyme moving along a DNA strand across different segments, or feeding rates of a predator over varying prey densities.
Median: The middle-most value. It is a positional average and is completely unaffected by outliers. Excellent for highly skewed biological data (e.g., survival time after a lethal viral dose).
Mode: The most frequent value. It is the only measure of central tendency that can be used for nominal/qualitative data (e.g., determining the most common phenotype or dominant allele in a field survey).
B. Measures of Dispersion
Central tendency only tells half the story. Evolutionary dynamics depend heavily on how data is scattered.
Range: Max value - Min value. Simple, but highly unstable because it depends entirely on two single data points.
Interquartile Range (IQR):
IQR = Q₃ - Q₁
Measures the spread of the middle 50% of the data, filtering out extreme outliers. Frequently used in box-and-whisker plots.
The average of the squared deviations from the mean.
Population Variance: σ² = Σ(Xᵢ - μ)² / N
Sample Variance: s² = Σ(Xᵢ - X̄)² / (n - 1)
The n-1 Correction (Bessel's Correction):
Why divide by n-1? A small sample tends to underestimate true variability. Dividing by n-1 mathematically corrects this bias, making s² an unbiased estimator of σ². Favorite concept for interview panels.
The square root of the variance. It brings the measure of dispersion back into the same units as the original data (e.g., changing cm² back to cm).
A dimensionless, relative measure of dispersion expressed as a percentage.
Why it matters: It allows you to compare variability between two completely different entities or units.
X̄ ± 1σ
68.27%
of the data
X̄ ± 2σ
95.45%
of the data
X̄ ± 3σ
99.73%
of the data
Scenario:
You are comparing the enzymatic activity of Catalase isolated from two different fish species acclimated to a temperature shock.
- Species A shows a Mean activity of 50 units with a Standard Deviation of 5.
- Species B shows a Mean activity of 200 units with a Standard Deviation of 10.
Question: Which species shows greater relative variability in its enzymatic response to temperature shock? How did you conclude this?
Hint: Use the Coefficient of Variation (C.V.) to make a fair comparison between these two groups.