Source

Adapted from my own handwritten and hand-typed notes from Year 1 to Year 3 of my EEE degree

Statistics uses a sample to learn about a population. The sample mean changes from sample to sample, so it is itself a random variable with its own distribution.

Population and sample

QuantityPopulation (parameter)Sample (statistic)
Mean
Variance

A sample must be random so that it is not biased towards one group (for example, sampling only students in a maths lecture is biased if the population is all students).

Distribution of the sample mean

For random samples of size from a population with mean and variance :

is called the standard error. Bigger samples give a less spread out .

When is X-bar normal?

Two cases:

  • if the population is normal, , then for any
  • Central Limit Theorem: if , then is approximately for any population

So we standardise with

Worked example

Students bring on average 1.75 pens, with standard deviation 0.65. For a sample of 50 students, find .

Solution:

  • , so by CLT

The matching result for the spread of a sample is in Sample Variance and Chi-Square Distribution.