Measures of Central Tendency — Meaning and Objectives

A measure of central tendency (commonly called an average) is a single value that attempts to describe, in one number, the general or typical level around which the individual values of a mass of data cluster. Instead of trying to grasp fifty individual wage figures at once, a single 'average wage' lets us summarise, compare, and reason about the whole distribution.

Characteristics of a good average (the requisites of an ideal measure of central tendency):

1. It should be rigidly (precisely) defined, so that different people applying the same definition to the same data always arrive at the same value.
2. It should be based on all the observations in the series, not just a few.
3. It should be easy to understand and easy to compute.
4. It should be capable of further algebraic treatment — for instance, the combined average of two groups should be derivable from their separate averages and sizes.
5. It should be least affected by sampling fluctuations — a small change in the data, or a different sample drawn from the same population, should not swing it wildly.
6. It should not be unduly affected by extreme (very high or very low) values. …