8. Sampling
Sampling
Method of collecting data about a large group by taking a sample
Sample: A small group within the large group (i.e. total population)
Used to make conclusions about the total population
Before sampling, need to decide on the interval (based on sample size) used to select members of the sample
Intervals can be in terms of:
Distance (e.g. every 2 metres along a coast)
Time (e.g. every 30 minutes at a traffic junction)
People (e.g. every fifth person who passes a particular place in the neighbourhood)
Two common types of sampling are systematic sampling and random sampling
Sampling (Systematic sampling)
Collecting data at regular intervals
Involves pre-selecting a number for the interval (e.g. 2 metres)
Useful when investigating changes e.g. along a river
Sampling (Systematic sampling)
Easy to carry out
But may miss out variations or changes in the total population biased data
E.g. By choosing equally spaced samples to study changes in a waterway, may miss the effect of canals joining the waterway
Sampling (Random sampling)
Collecting data at random or irregular intervals
The steps for random sampling are:
Decide how many samples needed
Obtain random numbers, either from a printed random number table or from an online random number generator
Use the random numbers to locate the samples
E.g.
Three water samples needed from a river
Random numbers are 20 and 17
Water sample would be collected at the 2-metre mark along the river, then at 1 metre after that (2 + 1), and 7 metres after that (2 + 1+ 7)
Water sample would be collected at 2 metres, 3 metres and 10 metres from the starting point
Best used when the population is the same throughout
Fair way of choosing (every member has the same chance of being picked)
Important to get a truly representative sample (same characteristics as the population) so that data obtained is accurate
Advantages
Saves time and energy
Essential because it is sometimes not possible to collect data about the total population
E.g. some people may decline to be surveyed
Disadvantages
May result in the data collection sites being widely spaced or concentrated in one area, leading to biased data
May not accurately represent the whole population