Representativeness is important because if your sample is representative of your target population then any results of your sample can be applied to your target population. this is otherwise known as generalisability.
As a general rule, sociologists must always try to make their sample representative. However, there are some instances where you might deliberately choose to study a non-representative sample. You might, for example, be interested in the social characteristics of the group itself, rather than what it may or may not represent. A type of unrepresentative sample is a best opportunity sample. This is where you choose a sample thst is best going to support your idea.
Also in some circumstances it may be imposibble to construct a representative sample for your target population. For example,if you wanted to study a secretive organisation such as a religious group that may refuse to disclose information to outsiders. If you do not know the social characteristics of the group itself, then you cannot know whether the sample is representative or not.
Examples: Goldthorpe and Lockwood-the affluent worker in the class structure. Best opportunity sample to prove that the working class was not becoming middle class.
- Electoral register
- School registers
- Professional membership lists - organisations such as the British Medical Association (BMA) keep a register of all doctors in Britain.
- Company Payrolls
They are very important because if the researcher cannot identify everyone in their target population it's unlikely that their sample will be representative and also, if the researcher is to make contact with the people in their sample they need to know clearly who the people are.
However, just because a list of a sample exists doen't mean the researcher has access to it. A school, for example, is unlikely to give an outside researcher access to their registers (legal reasons). A buisness organisation is unlikely to give an outside researcher access to their payroll records (confidentality). Some religious groups or political parties etc. do not want outside researchers to study their activities.