One of the most important aspects in any sampling design in research is the frame. The sampling frame has far-reaching implications in any research work in terms of cost and also the quality of research. A flawed sampling frame may result in undercoverage of important subgroups in the population under study.
A list of people or the items from which a sample was taken is referred to as a sampling frame. For example, if the workers in a factory are the population under study, then a single worker is the unit of that population. If all the factories in a given country are under study, then a single factory is the unit of the population. A sampling frame represents members in a population who can be sampled. The frame (data source) may include individuals, institutions or even households with details on their addresses, products produced by companies, the amount of their expenditure as well as their revenue data. An ideal sampling frame should be up to date (changes in contacts, addresses, fax numbers), have a complete list of all eligible geographical area to be covered, and should hold certain fields for each unit to facilitate stratification.
Actual sampling frames may deviate from ideal because of people moving in and out of a certain population target or even due to death. It is therefore not possible to have a fully up to date sampling frame (Ruoskokoski et al, 1998). For example, if we wish to carry out a research on the factors that lead to patients being admitted into the hospital following a malaria outbreak in a given area (population). Here the list of all the patients in the hospital is important for the research. From the list of all the names a convenient number is selected then (sample) as representatives of the population. This convenient number will participate in the research.
The sampling frame (a specified hospital) to be used should base completely on the population to be targeted. In the above example the sampling frame should not include patients outside the hospital since they are not in the population coverage. Inclusion of such members may alter the intended results of the survey. Accuracy in the sampling frame can be enhanced by ensuring that each member in the population sample is included only once.
The sampling frame is very critical in any research work as it saves on the costs that would otherwise be incurred in studying the whole population. The quality of survey is also enhanced by a sampling frame since the area of concentration is narrowed down and reduced in terms of numbers for greater efficiencies in research.
A poor sampling frame may hinder the research process in terms of expenditure. This means that a poorly constructed sampling frame may lead to huge budget in conducting the research that may not be practicable. The amount of time taken to carry out the research would also increase. Information derived from the research would be inadequate if the sampling frame is poorly constructed since the population would not be properly covered.
Convenience sampling involves selection of the subset of a population based on the researchers own judgment (the researcher decides whom to sample). For example, a researcher may apply this form of sampling by allowing only volunteers to participate in the research process. A researcher collecting data in a hospital may decide to select the first five names in the hospital’s register. In doing this the researcher excludes a great proportion of the population from the research.
Studies that employ convenience sampling are not invalid. This is because convenience sampling helps the researcher to conduct a pilot test in order to determine and obtain some basic information about the area of research prior to actual data collection. For example, choosing the first five patients from the hospital register may help the researcher to gain some insights on the expected outcome on the final research based on the data deduced from the patients selected. Convenience sampling helps in drawing conclusions that particular qualities occur in a given sample with great ease and with minimum time wastage since respondents are selected based on the researcher’s choice.
When using convenience sampling, however, it is important to state how the selected sample would differ from one that would be randomly selected. Description of the individuals who were left out during the research is important. This enables the readers of the research work to point out the difference between the actual results (from convenience sampling) and the results that would have been deduced from the entire population (Castillo, 2009).
Validity means soundness or solidness of a particular phenomenon under study. Validity in data collection means that the findings of a given study reflect the phenomenon that the research was measuring. Validity in research is important since it signifies solidity of the study. External validity refers to the extent to which the results of the research study can be applied (generalized) to other people in the sample. This means that the researcher can apply what he or she found in the study to other people.
Random sampling is a technique where all members in a population have an equal chance of being selected to participate in the study. In random sampling the units selected for the purpose of sampling are chosen by the use of probability methods and these help us to make a general perspective of the entire population (statistical inferences) from the sample of the entire population (Lund Research, 2010). Random sampling is therefore a true reflection of external validity. This is because generalization is possible using this method of sampling. Where questionnaires are used, for example, to representatives picked at random as opposed to where questionnaires are given to friends; a researcher is better placed to make a general perspective from the sample to the entire population.