Key Concepts

Reliability - This basically means consistency of results. It refers to the degree to which something, if repeated, would give the same or very similar results. It is usually assumed, that given a carefully chosen and representative sample and a well constructed questionnaire that a social survey would be reliable. Quantitative methods, by their very nature, are thought to be more reliable. If something is reliable it is more likely to lead to scientific generalisations.
Validity - This refers to the extent to which a research measures what it actually set out to measure, and the extent to which the findings are a true reflection of people's beliefs, attitudes and behaviour. It is usually assumed, again given carefully planned research, that qualitative methods such as unstructured interviews and participant observation produce valid data.
Representativeness - This refers to whether the group or situation being studied is typical of others. If this is true then we can assume that what is true of this group is also true of others and the work can be generalised. Careful sampling techniques are essential in ensuring representativeness. Quantitative data is more likely to be representative as it usually involves a big sample.