Quiz 14: Sampling
It becomes impossible for a researcher to collect data for all the candidates in a population. To study a population, the researcher selects a group of candidates through various sampling techniques to form a group that represents the population. The results obtained for the sample can be generalised for the entire population. a Parameters are the summary of all the descriptor variables of a population. They help in analysing the data of the entire population. Statistics are values summarising variables of a sample. They are calculated using the sample data which mimics the population. e.g. mean, variance etc. b A population refers to the entire collection of elements on which a study is required to be conducted while the sampling frame is a representative group selected from the population on which the experiment is conducted. c Unrestricted sampling is a method in which a sample is selected from a population without following any specific techniques. Restricted sampling is a method in which sample is selected using either probability or complex probability sampling techniques. d Simple random sample technique ensures that each of the sampling elements in a population has an equal chance and non-zero probability of being selected as a sample. Complex sampling can also be referred to as mixed sampling. The method includes both probability and non-probability sampling techniques to select a sample. e Sample selection done that is based on the convenience with each a candidate can be reached is known as convenience sampling. Sampling done based on a specific purpose of varied representation of the population is known as purposive sampling. f Sample precision refers to the closeness of values between samples while sample accuracy refers to the closeness of the value to the true value. g The variance that is contributed by the total variance of the variables under study by the researcher is known as a systematic variance. The variance in values produced from variables that are not under study is known as error variance. h Variables are elements whose value varies from one individual to another. Attribute parameters are different characteristics of a model that has specific values. i In proportionate sampling, the proportion of each stratum in a sample is equal while in the disproportionate sample, the proportion of each stratum is not the same.
Different sampling techniques can be used on basis of the conditions and budget of the study. Once a sampling technique is selected, it cannot be changed later. A correct sampling procedure is needed to develop meaningful conclusions regarding the population and reduce errors. a Probability sampling is used in studies when high precision is required in the sample to represent the population. The researcher can make probability-based estimates of various parameters confidence intervals. Non-probability sampling is followed when the cross-section of the population need not be accurate. Used in cases where the study needs to be done in low budget or sampling does not play a role in variable measurements.. b Simple random sampling is the most basic and easiest form of sampling technique. Each element has an equal non-zero known chance to get selected in the sample. The sample size selected through this can be larger than the size of the sample selected by other methods. Hence, it can be expensive and time consuming. In cluster sampling, a population is divided into clusters of unequal size and some clusters are randomly selected as sample. This is followed either to decrease costs or when probability sampling cannot be performed. In stratified sampling technique, a sample is divided into multiple strata. Elements are selected randomly from each strata to form a sample. This technique is followed to increase sample's efficiency, provide precision data and enable research methods to be used for multiple strata. c When selection from each strata is not done proportionalely then the sampling is known as disproportionate stratified sampling. It is followed when each strata is large enough to provide required confidence interval. Also, when costs of sampling among strata are unequal.
If an element is selected from a population individually, then it is known as unrestricted sampling. No complex sampling rules are applied in unrestricted sampling. To conduct a survey using unrestricted sampling, one can either select random probability or convenience sampling. The subjective decisions to be taken by the researcher depends on the sample precision required and the time available. • Probability sampling method randomly selects elements from the population. It is easier to implement but the sample size required to attain precision is usually large. • In convenience sampling, the researcher has the liberty to choose anyone based on judgement. This is the least reliable but easiest and the fastest method.