In this method, the population is divided into strata, such as social classes or occupational groups, and a sample is randomly selected from within a particular strata. This method allows the researcher to focus on a particular group, such as women or women who are also managers, but it is important not to allow the strata to get so small that they cease to be representative.
A quota is decided on the basis of characteristics of the population, such as age group or gender, and no attempt is made to select at random. For example, the sample may comprise all men aged 65 years or over. This is a method which is widely used in market research, but is not suitable for academic studies because bias can easily be introduced.