Populations cannot be homogenous, that is, elements of the population cannot be identical. Some populations may contain variables that are closer in similarity, though most often this is not the case. To counter this, probability sampling eliminates “systematic bias, the over- or under-representation of some population characteristics due to the method used” (Schutt, 2012: pp. 144). This method strives to select a random sample that is most representative of the population as a whole. In doing so, larger populations allow for more consistency in sampling a division that is characteristic of the totality and therefore, lead to better generalizations when relating findings back to the population. Neither small nor large populations will lead to exact findings, however, smaller populations may incur less of a chance the sample will embody the whole …show more content…
These can be obtained in many ways and are usually less complicated to setup than probability samplings. Non-probability samplings can be broadly divided into the categories of accidental and purposive. One example of accidental, or haphazard sampling, would be standing on a street corner and conducting interviews. This would appear to give a quick ‘read’ of public opinion, but in reality is not representative of an entire population. Purposive, non-probability techniques are set up with a specific determination in mind. In other words, the sampling is based on a preconceived indication of the expected outcome. An example of this would be expert sampling – knowingly eliciting the views of individuals who have a specific skillset or a specific type of knowledge. While non-probability techniques do have the advantage of being less expensive and easier to obtain, they do not give a proportional representation of the population and would not yield reliable generalizations. These non-random samples are best suited as an exploratory method for obtaining some initial insight, perhaps as a prelude to a main study. (Schutt, 2012: pp.