In research, population (universe) is collection of entire elements that researchers want to investigate. The data set of CBS is not only the population and is determined by area of research problem. For example, if researchers want to conduct a research on the effects of IT in college going students of Kathmandu valley than all the students who joined college in Kathmandu valley is considered as population in research. Thus, it is collections of people or objects which researcher want investigation in systematically way.
The research population may be finite or infinite. If the elements of population can be determined than it is called finite population. The population of study on the psychological behaviour of bachelor level rural development of 2076 batch of K&K international college is finite population because researcher can easily find out total admitted student of rural development in K&K international college in 2076 from internal source. If researcher only able to determine the partial elements but unable to determine the entire elements through any means than it is called infinite population. If researchers want to study on the psychological effects on women who suffers from domestic violence of Kathmandu distinct that researcher does not able to find total elements. It is an infinite population in research. The definition of finite and infinite population in research dependent on how broadly or narrowly researcher define population and availability of secondary related to population.
If researcher selected partial element from population for investigation than it is called sample. Thus, the sample is the collection of partial population element which researcher believed that it is true representation of the population under investigation. Sample is always smaller than population in size but contain almost the same characteristics of population.
The process of selecting sample for investigation from the population is called sampling. Research may use probability or non probability sampling methods for sample selection. The choice of methods of sampling is determined by various factors like, types of research, philosophy of research, available population element, objectives of investigation and so on. Researcher must carefully use the sampling method for conducting research because wrong sample selection manipulate the collected data and there is the question of epistemological, ontological and methodological consideration while generating knowledge. If researchers collect data from the entire population than it is called census method which is far better method than sampling method. But there are many constraints for performing the investigation from census methods. Availability of Time and cost incur are among the major constants for using census method in research. Besides that complexity of human behaviour, difficulties of analysis of data are next to them.
For selecting sample, at first researcher define and identify the population under investigation. After identify the population researcher must develop sampling frame. Sampling frame is the identification code of population element. Researcher than choose the appropriate sampling method for selecting sample. The choice of method is dependent on the types of population, nature of sample, qualitative or quantitative research and so on. After choosing the sample method the researcher determine the sample size using appropriate statistical tool. Researcher than select sample and collect data from sample.
After identification and defining population researcher must develop sample frame for the selection of sample. Sampling frame is the list of all the items of population and use for identification of elements for sample. It provides a complete list of everyone or everything of population under consideration. In sampling frame researcher give identification code, number or elements of population which makes easy to select the elements for sample.
While developing the sampling frame researcher consider the set of elements for selecting sample. For example, researchers may use mobile number for individual people, household number for selection of HHs, class roll number or exam symbol number for selection of students and so on which are act as identification code for selecting sample and these are used in sampling frame. If a researcher does not able to identify unique element in sample units than give some arbitrary code for identification but normally, there is no need of all these things.
While developing the sampling frame researcher must consider following possible sampling frame error (Lessler & Kalsbeek, 1992):
- Inclusion of units not in sample (over coverage)
- Units listed more than one (duplication)
- Units from dissimilar clusters (possible unequal weighting)
- Missing sample units in frame (under coverage)
- Units for use in satisfied sample design (missing unit information)
- Frame list out of date (sample does not represent population)