Systematic sampling
Systematic sampling selects every k-th unit from a numbered population list after one random starting point, spreading the sample evenly without drawing k separate random numbers.
See it move
A retail chain numbers 900 customer accounts and wants a sample of 45, giving a sampling interval of 900 ÷ 45, or 20. A random start of 7 is drawn, so the sample runs 7, 27, 47, 67, and onward, adding 20 each time until 45 accounts are selected.
The formula
Variables
- Sampling interval
- Total population size (units)
- Desired sample size (units)
Round k down to a whole number if it does not divide evenly, then draw the random start from 1 to k.
Check yourself
A researcher wants a systematic sample of 60 invoices from a population of 1,200 invoices numbered sequentially from 1 to 1,200. She draws a random starting point of 9. What is the number of the fourth invoice selected in her sample?