Contents
Meaning of Sampling
Sampling refers to the method of selecting a small pattern of data from large population for the purpose of carrying out an investigation. The selected pattern is termed as sample which is a small and manageable version of large set of data. Sampling is most widely used in statistical testing where size of population is too large such that it is impossible to include each individual observation in test.
Under this technique, to ease the process of doing a research on whole population, it is divided into small sampling unit. These sampling units represent the characteristics of whole population and should not reflect bias towards a particular attribute. Samples drawn from population are used by researcher for making statistical inferences and estimating the information about whole population. Methodology to be used for the technique of sampling depends upon type of analysis being conducted by researcher. Probability and non-probability sampling are two common sampling methodologies. Sampling is mostly used by businesses for studying the needs and preferences of people in market.
Characteristics of Sampling
Various characteristics of sampling are discussed in points given below: –
- Goal-oriented: Design of sampling should be goal oriented. It must align clearly with the objectives of research being conducted and should be in accordance with conditions of survey.
- Proper universe representation: Sample chosen should adequately represent the characteristics of whole population from which it is taken. It should fairly represent details about all units without any biasness. There are different methods of choosing a sample and it need to be chosen with utmost care as improper sampling would lead to error in survey.
- Proportional: Size of sample should be proportional with the size of population. It should be large enough for representing the whole universe and must provide statistical reliability. Sample must ensure proper accuracy for carrying out the particular research study.
- Economical: Process of sampling should be economical requiring minimum cost and efforts for attaining the objectives of survey.
- Random selection: Sample units should be selected on a random basis under which every unit has an equal chance of being chosen. It will ensure that sample is a fair representative of whole population.
- Practical: Design of sample should be simple and practical. It must be capable of easily understood and applicable in fieldwork.
Types of Sampling
Various types of sampling are as discussed below: –
- Random sampling: Random sampling is a technique under which every member of population has equal chance of being selected in sample units. It is most reliable method which ensures fairness and eliminates any biasness. Under random sampling, whole population need to be properly numbered or names should be allotted to it and then a raffle method is used for making the sample.
- Convenience sampling: It is a technique under which individual from target population is chosen on the basis of their easy availability and willingness to take part in survey. Convenience sampling is an easy and inexpensive method under which participants are chosen by researcher on the basis of their easy accessibility. However, this method may not represent whole population accurately and involve biasness.
- Systematic sampling: Systematic sampling is method in which participants are selected from population using a systematic/orderly manner. All members are properly numbered and then chosen at regular intervals instead of randomly generating numbers. This sampling technique is less time-consuming as it has predefined range.
- Stratified sampling: Stratified sampling is a type of sampling under which whole population is divided into distinct small sub-groups based on various individual traits such as gender, age, job role and income. Groups are formed in such a way that it does not overlap. Peoples in each sub-group are included on the basis of overall proportion of population.
- Judgmental or Purposive sampling: Under this type of sampling, judgements of researcher is used for choosing sample units. It is also termed as selective sampling in which samples are formed at the discretion of researcher.
Advantages of Sampling
Various advantages of sampling are as discussed below: –
- Lower sampling cost: Sampling reduces the overall cost involved in doing research. The cost for collecting data about entire population is quite high. Sampling reduces the population into small manageable units. Acquiring data about sample of population involves lower cost which is one of the major advantage.
- Less time consuming: Sampling reduces the overall time by reducing the size of population. Data is not collected about every member in population but only related to sample is gathered. It is less time-consuming in comparison to census technique.
- Higher accuracy of data: A sample represents the whole population from which it is drawn. It is used for calculation of desired descriptive statistics and a stability of derived sample value can be easily determined. Samples permit a high level of accuracy because of limited area of operations. It enables in proper execution of field work and results of studies conducted on the basis of theses sample units turn out to be accurate.
- Higher scope of sampling: Sampling enables investigators to easily arrive at generalizations about set of data. It would be totally impractical to study whole population as it is too large for measuring characteristics of all individual members. Process of sampling by analyzing variables within small proportion of population ease in arriving at generalizations.
- Intensive and exhaustive data: In studies based on sample units, observations are made of a limited number. Therefore, exhaustive and intensive data are collected.
- Suitable in case of limited resources: Sampling is very effective technique of collecting information in presence of limited resources with organization. Studying the whole population requires large amount of resources both in term of money and time. Sampling makes it possible to cover whole population satisfactorily even by employing limited resources.
- Better rapport: Good rapport in between the researcher and respondents is must for carrying out an effective research study. In presence of large population, various issues of rapport arise.
Disadvantages of Sampling
Accuracy of sample is dependent upon appropriateness of sample method used. Theory of sampling focuses on improving the efficiency of sampling. Major difficulties are pose at the time of estimation, selection and administration of samples. Various disadvantages of sampling process are discussed in points given below: –
- Chance of Bias: Major limitation that arises with sampling is chance of biasness in choosing sample units. Selection of samples is a judgmental task as it is based on mindset of individual choosing them. These biased selection does not truly represent the whole population and may lead to faulty conclusions by researcher.
- Difficulty in choosing a truly representative sample: Choosing an adequate and reliable sample that is a truly representative of population remains a difficult task. In case the phenomena under study is of complex nature involving heterogeneous data, it becomes difficult to select proper samples.
- Lack of adequate subject knowledge: Application of sampling process requires proper knowledge regarding sampling technique by individual selecting sample units. This process requires computation of probable error and statistical analysis. There are chances of serious mistakes being committed by researcher in case if he lacks specialized knowledge about sampling. Consequently, overall results of research study conducted will be misleading.
- Impossibility of sampling: Process of sampling is not applicable in cases where universe is too small consisting of heterogeneous set of data. It is difficult to derive a representative sample in such cases. Census study is the only alternative for doing study for such phenomena. Also, sampling is inadequate for studies that needs a high degree of accuracy. There are always chance of errors in sampling even if sample units are chosen with utmost care.
