Measures of Central Tendency

Measures of central tendency are single values that summarise where the bulk of a dataset lies — the typical or central observation around which the others cluster. The three classical measures are the mean, the median, and the mode, and choosing among them depends on the level of measurement and the shape of the distribution.

Definition

A measure of central tendency is a single value that identifies the centre of a distribution: the arithmetic mean is the sum of values divided by their count, the median is the middle value when observations are ordered, and the mode is the most frequently occurring value.

Scope

This entry covers the mean, median, and mode: how each is computed, what it represents, and when each is the appropriate summary of location. It is a methodological reference and does not provide clinical guidance.

Core questions

• Which measure of location best represents this variable?
• How does distribution shape affect the choice between mean and median?
• When is the mode the most informative summary?

Key concepts

• Arithmetic mean
• Median
• Mode
• Robustness to outliers
• Effect of skewness on the mean and median
• Measurement level and choice of average

Mechanisms

The mean uses every observation and is the natural summary for symmetric, interval- or ratio-scaled data, but precisely because it incorporates all values it is pulled toward extreme observations and is distorted by skew and outliers. The median, the middle value of the ordered data, ignores the magnitude of extremes and is therefore robust, making it the preferred summary for skewed continuous data and ordinal variables. The mode, the most common value, is the only measure applicable to nominal data and is useful for identifying the most typical category or a peak in the distribution. In a perfectly symmetric unimodal distribution the three coincide; as skew increases, the mean is displaced furthest in the direction of the tail.

Clinical relevance

Reported averages — mean blood pressure, median survival, the most common diagnosis — are central to how clinical findings are communicated, and recognising which measure was used guards against misreading skewed data. This entry describes how location is summarised for appraisal and is not a basis for individual diagnostic or treatment decisions.

Epidemiology

Because many health measurements are skewed, the median is frequently the more faithful summary of a typical value, and reporting a mean for such data can overstate the central value. The choice of measure therefore affects how population characteristics and outcomes are conveyed.

History

The arithmetic mean has been used since antiquity for combining measurements, and the formal distinction among mean, median, and mode was consolidated as descriptive statistics matured in the nineteenth and early twentieth centuries. The recognition that the median better represents skewed distributions is a long-standing principle reiterated throughout the applied statistical literature.

Debates

Mean or median for skewed clinical data?

For right-skewed quantities common in medicine — costs, lengths of stay, biomarker levels — the mean is inflated by the tail while the median tracks the typical value, so guidance generally favours the median, with the mean reserved for roughly symmetric data.

Key figures

• S. Manikandan

Seminal works

• manikandan-2011-mean
• manikandan-2011-median-mode

Frequently asked questions

When should the median be reported instead of the mean?

When the distribution is skewed or contains outliers, or when the variable is ordinal. In those situations the median represents the typical value more faithfully than the mean, which is pulled toward the extremes.

Can the mode be used for any kind of data?

Yes. The mode is the only measure of central tendency that applies to nominal (categorical) data, and it can also highlight peaks or the most common value in numerical data.