Choosing the sample population
Who are you going to interview?
A population is every possible person who could be interviewed for your research (David and Sutton 2004). For example, in a study of first year student debt amongst full time students at the University of the West of England, Bristol, the population would be all first year, full time, students studying at the university. Of course, the individual units within the population do not have to be people, they could be schools, businesses or households etc.
Often it will not be possible to carry out a census, and so a sample or representative group will need to be selected from your population. This is an important stage of your research as you may want to use the sample data to draw inferences about the population as a whole.
Sampling is a term that refers to the selection of a subset of individuals from a population who will form the 'sample' of the population that you will interview. There are 2 main types of sampling technique; probability sampling and non-probability sampling. With probability sampling techniques, you randomly select a sample from the population. This means that you have to know who your population is and have a comprehensive list of this population (a sample frame). Probability sampling techniques are compatible with statistical analysis. Non-probability techniques can be used when you do not know who your population is or you have no sampling frame. These techniques tend to be simpler than probability techniques. However, they may not be suitable if you wish to carry out statistical analysis. Non-probability sampling is also more prone to bias.
The kind of sampling that you undertake on your population depends upon the type of interview you will be conducting. For structured and semi-structured interviews, random sampling methods are most often used (probability sampling techniques). For unstructured interviews, snowball or opportunity sampling methods are more suitable (which are non-probability sampling techniques).
Random sampling (for use with structured and semi-structured interviews)
Random sampling involves selecting individuals randomly from the sample frame. The simplest form of random sampling would be to put all the names in a hat and draw them (whilst wearing a blindfold to avoid bias!). Generally, random selection is now carried out by computer.
Snowball sampling (for use with unstructured interviews)
Here you may find one person to undertake your survey and ask them to recommend other people that you could also interview. If you have a small or rare population, or do not know who your population is e.g. hard drug users in your city, this may be the only way to achieve a sample.
Opportunity sampling (for use with unstructured interviews)
Opportunity sampling involves interviewing people who happen to be available at the time of the study and who meet your criteria. For example if you were carrying out research into student debt amongst first year students at the University of the West of England, you might attend the students union at the start of term and interview those 1st year students who happen to be there.