When a study is being conducted on primary data, a frequent dilemma that is faced by almost all the researchers is that of choosing the right sample size. A student of postgraduate class is undertaking a research as a part of his curriculum, with a packaged juice manufacturing company. His project is to study the popularity of packaged juices among the young residents of the city in the age of 18-30.

While deciding on the sample size, the researcher decided to use a non- statistical approach. He got advice from all sides. Some said that refer to similar studies conducted earlier and adopt the sample size chosen by them. Others suggested that he go for a typically used size e.g. of 100 respondents. It was like picking up a magic number, which is a multiple of 5.

The company with which he was doing his training suggested that he chose a sample size tat fits with the budget being provided by the company for conducting the research e.g. the company was giving an amount of Rs. 10,000. The supervisor provided him with the following estimate:
Total amount available                                        Rs. 10000
Minus fixed costs                                                      4000
--------------         
Amount available for variable cost                              6000
Divide by the cost of contacting each respondent            50
--------------
120
--------------
Hence a total of 120 respondents were suggested as the sample size.

Another option was to study the demographic profile of the population and decide on the criteria against which the analysis is to be carried out e.g. the population can be categorized into two major sub- groups of males and females and then a minimum of 100 respondents from each of these subgroups occupation (e.g. student, government service, private sector, business, homemaker)  it has resulted in five relatively smaller subgroups . In such situation 50 respondents from each group can be chosen. This will result in a sample size of 250 respondents.
Thus there are a plethora of criteria available to the researcher for deciding the optimal size for his project. Yet life was never more complex.