What is the difference between a select and random sample?
In a previous blog, I wrote about how best to draw a sample. In the blog, I briefly described the difference between a select and a random sample. There is a substantial difference between the two.
In a random sample, everyone in the focus population has an equal chance of being in the sample. This sample is also called a probability sample or random selection. There are several methods for doing a random sample:
- Single random sampling: the lottery method, you randomly select a number of people. This can be done manually or by computer.
- Systematic random sampling: you choose a random number and then go through your list in steps, for example, 20-120-220 etc.
- The cluster or bunch sampling: the list is divided into several clusters. And then, a group is chosen at random.
- The quota sampling: In advance, it is known how many respondents are needed; when that number is reached, the data collection stops.
In a select sample, not everyone has a chance to be in the sample group. The results apply only to the group being studied. There are a number of selective sampling options:
- Snowball sampling: At first, one person is talked to; this person is asked if they know anyone who would also be interesting to talk to. And so on until the sample is large enough. When doing this, make sure you approach people with different perspectives; otherwise, your sample is flawed. Use this method when conducting interviews.
- The convenience sample: the researcher approaches people from his circle or the circle of colleagues until the sample is large enough. Of course, this sampling method is not fully representative because the respondents are chosen on insignificant factors (would know).
- Stratified sample: you divide the population into different groups. A representative number of respondents are chosen from each group.
- The two or multi-stage sampling: a sample is first drawn from several main categories. After this, a representative number of respondents are chosen from the selected main categories. For example, first a number of cities are determined, and from these cities, a neighborhood and then a street per neighborhood. This method is not a good reflection of the population.