# What are the primary strengths and weaknesses of simple random samples, systematic samples, stratified random samples, cluster samples, quota samples, convenient samples, purposive samples...

What are the primary strengths and weaknesses of simple random samples, systematic samples, stratified random samples, cluster samples, quota samples, convenient samples, purposive samples (judgmental samples), and snowball samples?

pohnpei397 | Certified Educator

Simple Random Sample

The main strength of this type of sample is that it will be highly representative of the population as a whole.  You will have a truly random sample which, by definition, should have similar characteristics to those of the whole population.

The main weakness is that it is very time consuming.  You have to assign a number to each individual in a population and then use a random number generator to select your subjects.

Systematic Sample

The main strength is that you still get a fairly high degree of representativeness without having to be quite as thorough and without having to take quite as much time as in a simple random sample.

The main weakness is that this is not quite as random as a simple random sample because you are taking every nth person rather than a truly random sampling.

Stratified Random Sample

The main strength of this technique is that it makes sure that you get adequate numbers from each of a set of groups that is important in your population.  It allows you to be sure, for example, that you get enough representatives from various races or income levels in a sample.

The main weakness is that this is at least as time consuming as a simple random sample.  This is true because you have to split your population into groups and then randomize within those groups.

Cluster Sample

The main strength of this technique is that it is relatively simple and easy.  When the population you want to study is made up of groups, rather than individuals, you can pick random groups and sample from them.  For example, since there is no list of all high school students in the US, you could randomly pick high schools (groups) and sample from them.  This is relatively easier than trying to find the identities of your entire population.

The main weakness is that the groups may not be representative.  You might, in our previous example, select non-representative high schools.

Quota Sample

The strength of a quota sample is that you can make sure that you get the right number of subjects from each group that you are interested in.

The main weakness of this sort of a sample is that it is not random.  Therefore, we cannot know for certain that we can really extrapolate our results to the population as a whole.

Convenience Sample

As the name implies, the main strength of a convenience sample is that it is convenient.  The researcher uses subjects who are at hand, which is much easier than going out and contacting random subjects.

The major weakness, of course, is that it is very hard to know if results will be generalizable to the entire population.  The sample is in no way random.

Purposive (or Judgement) Sample

The strength of this kind of sample is that it allows the researcher to think about the sample they want based on their expertise on the issue.  The researcher can pick the people whose opinions will shed most light on the issues of interest.

The major weakness of this sample is that it is in no way random.  It may well be skewed by the researcher’s own biases and blind spots.

Snowball Sample

The major strength of this kind of sample is that it allows researchers to find members of populations that are not easily identified.  For example, I once worked with a professor on a project where we used a snowball sample to find lawyers who had a reputation for being involved in issues of social justice.  There is no easy way to find such lawyers, so we started with ones we had heard of and then asked them who we should interview.  This allowed us to find good interview subjects from a hard to identify group.

This is another type of sampling that is weak due to its lack of randomness. It can be strongly affected by the biases of the people who identify possible subjects.