What are the disadvantages of probability sampling?

Answer 1

The three main disadvantages are non-response, representation and expense.

Under probability sampling once the sample is selected the researcher may not interfere with it sample. Specifically, if a large proportion of the sample do not respond, the study will be seriously flawed. With quota sampling, for example, the sampling will continue until the required number of responses are obtained.

The representation issue arises if a small sample is taken from a large population which is made of of sub-populations. Elements within each sub-population are similar to one another but there are important differences between members of different sub-populations. Some sub-populations may end up not being represented in the sample. A typical school-level example is sampling all employees of a company: 2 managers, 90 workers, 8 office staff. If using 10% samples, most (96% of them) will exclude any managers. Stratified sampling will get around that.

The expenses issue can be best illustrated by an example. Suppose you want to sample 1% of all schools in your country. You get a list of all schools in the country and select a 1% random sample. These will be scattered all over the country and you (or your team) will spend a huge amount of time and money travelling from one location to another. A better alternative is to divide up the country into a number of regions, each with [approximately] the same number of schools. Select a 1% sample of the regions [not schools] and then conduct a census of all schools within the selected regions. A variant might be to sample 2% of the regions and, within each region sample 50% of the schools so as to give a 2%*50% = 1% sample overall.