Random and non random …that’s it!!

In the last post, we talked about 2 types of sampling after which you were to find out at least 2 major differences between the two.

My sole purpose for this task was to get you to think in the lines of the 2 broad categories of sampling.

Most of you have reasoned perfectly, saying that “simple random sampling” allows everyone a chance of getting selected whereas “convenient sampling” doesn’t allow an equal chance to every one and is biased, some of you have also said that one is random and the other is non random. Consolidating what all of you have said there were 2 major differences that were found;

• One was the randomness of the sampling techniques - simple random sampling allowed every person in the sampling frame an equal chance of getting selected in the sample, whereas convenience sampling gave an unequal chance to the persons in the sampling frame or rather it did not have a sampling frame to start with. Thus one allowed random sampling whereas the other was non random sampling. On this very basis stand the two broad types of sampling,
  
  
  

  - Random sampling
  
  - Non random sampling

Other names for the 2 are

Probability sampling (Random sampling) – this name because it allows
an equal
probability (chance) to the individuals/items in the sampling
frame. There are
many more sub-types of which we will consider 6, out of the
6 we have already
considered simple random sampling, the others are:

1. Systematic sampling


2. Stratified sampling


3. Cluster sampling


4. Stage sampling


5. Multi- stage sampling

Non probability sampling (Non random sampling) this name because it allows an unequal probability (chance) to the individuals/items in the sampling frame. It also has many sub-types out which we will deal with 5, one of which is convenience sampling. The others are :

1. Quota sampling


2. Purposive sampling


3. Dimensional sampling


4. Snowball sampling

• Other was the generalisability issues of the sampling techniques: this issue stems from the probability of getting selected in the sample. Random sampling allows equal chance or probability to all individuals hence the sample is free of any bias, since it is free from bias it represents the population from which it is selected, and hence the results are generalisable to the sampling frame or population. Non probability sampling on the other hand does not give equal chance of getting selected in the sample to all the individuals in the population; instead it selects the people/items easily accessible to the researcher. Hence there is an element of bias introduced in it, as preference is given to those people who are easily accessible and not to those without easy access. Thus the results of a study involving such a sample will be applicable only to that sample and not the entire population.
  
  Click on the following image to look at what we just discussed in a more visual form: