Yes...randomness has types!

As mentioned in the earlier post the 2 examples show us how randomness can be achieved. I had asked for the differences in which the participants were selected. Most of you have written what I expected. The differences are as follows:

1. In example 1, the researcher randomly chooses a sample of 300 teachers from a population of 1000 teachers. This is called random selection. In this case the sample of 300 teachers is representative of the population of 1000 teachers. Thus the results obtained from these participants will be applicable to the larger population (in this case, the population of 1000 teachers). Hereafter the participants are not segregated into different groups. They are treated as one big group i.e. sample which is the requirement of the study. Thus we can say that in random selection the individuals are selected randomly as representing a population. You may have a question as to how these 300 teachers were selected; they can be selected from the many types of probability sampling that you will discover in the next post. In random selection we can be sure that the results will be genralisable to the population.


2. In example 2, the teacher decides to select 50 students and then segregates them in to 2 separate groups required for her study. She gives each of those 50 students an equal chance of being assigned to either one of the groups thereby maintaining the randomness in which the students are grouped. Such kind of random assignment is common in experimental studies. Since in this example, the participants need to be divided into groups for the study, a question arises as to who should go in which group. This question is answered by the method in which the teacher divides the participants into groups. Any other method would have brought in bias and reduced the randomness. This method of achieving randomness is called random assignment.

Thus in this post we’re introduced to 2 methods in which randomness is achieved
· Random selection (if you cannot see the picture clearly, click on it)

· Random assignment (if you cannot see the picture clearly, click on it)

It should however be noted that in random assignment the method in which the intact group (the group of participants that has to be segregated for different treatments) is selected could influence aspects such as generalisability and representativeness towards the larger population. It could happen that the intact group could be selected using random or non-random sampling. Hence if the intact group is selected using random/probability sampling, then the results obtained from the divided groups of participants will be applicable to the population from which it was selected. Otherwise the results will be applicable only to the intact group and not the entire population.

Another aspect of concern in random assignment is that if the number of participants in the intact group is not equal to the number required, then the excess number of individuals are eliminated at random. Meaning, in the 2nd example if the number of students in the teacher’s batch is 55 instead of 50 then 5 students will randomly be selected from those 50 students to be excluded from the study.