The quest for universally applicable principles is the main goal of research. The majority of studies apply treatments to a limited subset of the population and gather data for analysis and conclusion-making. There are several pragmatic reasons for doing this. Certain populations can be so big that it would be prohibitively costly and time-consuming to quantify their features. When dealing with extremely huge populations—such as all Pakistani primary school students or all automobiles registered in the nation—the process of gathering data may take so long that, by the time the researcher finishes his investigation, the population may have changed. If such were the case, the findings from the research would not apply to the altered population, whose makeup may have been impacted by a higher percentage of new enrollees and a smaller percentage of repeaters and dropouts.
In certain situations, it might seem possible to gather information by conducting an exhaustive census of the entire population, but in the majority of other cases, sampling would allow one to draw reliable conclusions or generalizations from the data gathered from a small percentage of the population. The sample size’s sufficiency determines how accurately we can make these Quake generalizations or inferences. The researcher is not the only one who uses sampling extensively. Everybody uses sampling in their daily lives. We remove one or two peanuts from the seller, remove the shell to check whether the nuts seem healthy, and then taste the peanuts to determine if they have been cooked to our satisfaction. We depart if the quality of the contents of these sampled peanuts does not meet our standards.
Likewise, our perceptions of organizations and people are frequently shaped by snippets of our interactions with them. We might not always make purposeful attempts to get high-quality samples, thus the conclusions we make in daily life based on samples may not always be accurate. Conversely, the researcher uses the following sampling strategies to create samples:
• randomly selected;
• representative;
• sufficiently large; and
• controlled for extraneous variables.
Read: Classification of tests: the validity and reliability of tests as research tools
Type of Samples
Non-probability and probability are the two fundamental categories of samples. What separates the two is primarily that in the former case, it is impossible to calculate the likelihood that every person has an equal chance of joining the sample. To clarify the distinction, consider the example that follows. Let’s say a researcher visits elementary school principals to gather information from those that allow the headmasters to provide their pupils for study purposes. He visits many schools, gathering information until he reaches his 250-point goal. This one is an illustration of a non-probability sample. In contrast, g gathers a comprehensive list of all class V students enrolled in schools, writes each student’s name on a separate slip, places all the slips in a drum or container, and then draws 250 slips, one at a time, after repeatedly rotating or shaking the container. This allows him to create a probability sample. Because each student has an equal chance of being included in the sample, this is known as a probability sample.
Non-probability sampling
The majority of social science research uses non-probability-type samples. Subjects are typically used when they are easily accessible to the researcher. For instance, a researcher may enroll in two distinct courses if they are interested in comparing two distinct pedagogical approaches. We refer to these samples as inadvertent or accidental samples. Entering a sizable department shop, an investigator may sample the prices of necessities by taking note of the costs of every item on offer. Incidental or inadvertent samples are used by a drug examiner who is gathering samples of medications accessible in a large pharmacy.
Quotation sampling is another kind of non-probability sampling. Using this procedure, the proportions of the population’s different subgroups are first discovered, and the sample is then selected to have the same percentages (typically not in a random manner). For instance, a researcher examining the mathematical prowess of high school students includes both males and females in the fraction of students known to be enrolled in high schools. Comparably, a researcher may take into account the proportion of urban and rural instructors in the population when examining the attitudes of high school teachers about the removal of the ban on student unions.
Purposive sampling is the third kind of non-probability sampling in research. A sample that is purposefully chosen is chosen at random because there is strong evidence supporting its high level of population representation. Given that the candidate with the highest number of votes in the Islamabad Capital Territory’s rural regions has historically won the National Assembly seat, anybody attempting to forecast the outcome of the next election from this constituency may choose to sample solely the voters’ preferences in this area. Alternatively, if one is aware that the Faisalabad wholesale market has historically mirrored national trends, one may use the trend in cotton prices there alone to forecast the trend in cotton prices for all other markets in the nation.
Probability sampling
A planned and systematic process of selecting samples from the population so that every member of the population has an equal chance of being included in the sample is known as probability sampling. For instance, if a sample of two students is to be taken from a class of ten students (S1. S2, S3, S4, S5, S6, S7, S8, S9, S10), one can create five pairs of students so that each student appears only once in a pair. Five pairs can be combined in a wide variety of ways, and any pair can be chosen.
Read our blog: AABMS Blogger
