# Sampling Definitions

Response Post 1:

1)Differentiate between probability and non-probability sampling.

As defined by Rubin & Babbie (2017, p.350), probability sampling involves random sampling. It includes specific, precise, and scientific sampling techniques that can allow for determination and control of the likelihood that particular elements are selected for the population being studied.

Conversely, non-probability sampling is used when it may not be appropriate or possible for the sampling to be selected randomly (Rubin & Babbie, 2017, p. 352). The sample chosen will not be one where everyone in the population of the study will have an equal chance of being selected to participate in the study (Radey, 2020).

Examples of non-probability sampling are (Rubin & Babbie, 2017, pp.353-356):

– Reliance on Available Subjects: also known as availability sampling, employs the use of sampling whoever is available.

-Purposive or Judgmental Sampling: selecting a sample based on your own knowledge of the population and purpose of the study.

-Quota Sampling: addresses the issue of representativeness by creating a matrix that captures the characteristics of each element of the population in the study. The elements are weighted so that the overall data will provide a reasonable representation of the total population.

-Snowball Sampling- collecting data from a few members of a target population and asking those members to refer others that they may know.

2)Discuss the logic of probability sampling.

As all members of a population are never going to be identical in all characteristics, probability sampling provides for an efficient method for the selection of a sample that will take into account variations of a population. (Rubin & Babbie, 2017, p. 357).

There are several ways in which a researcher can encounter obstacles in selecting probability sampling. One risk inherent to probability sampling is conscious and unconscious sampling bias (Rubin & Babbie, 2017, pp. 357 – 358). Choosing participants can be affected by feelings of the researcher of which they may or may not be aware. Those selected for participation are not typical or representative of the broader population.

Another piece on the logic of probability sampling is representativeness and probability of selection (Rubin & Babbie, 2017, pp. 358-359). A fundamental principle of probability sampling is that the sample will be representative of its population only if all members of the population have an equal chance for selection. Probability sampling offers two advantages in the areas of representativeness and probability by taking into account and avoiding bias as well as using probability theory to estimate a sample’s accuracy.

The final piece used in the logic of probability sampling is the use of random selection (Rubin & Babbie, 2017, p.359-360). Key to the process of probability sampling, random selection allows each element to have an equal chance for selection independent of any other event in the selection process. Random selection serves to alleviate conscious or unconscious bias and allows the use of probability theory for estimating the characteristics of a population and the accuracy of samples.

3)Give descriptions of the basic types of probability sampling – simple random, systematic, stratified, and cluster.

Simple random sampling occurs when the researcher assigns a number to each element of the sampling frame. A table of random numbers or a computerized random number generator can then be used to select elements for the sample (Rubin & Babbie, 2017, p. 366).

Systematic sampling is more complicated but is more efficient to use when a population list is available. (Rubin & Babbie, 2017, p. 366). As Radey (2020) explains, in systematic sampling, every kth element in the list is selected. The distance between the elements chosen for the sample is known as the sampling interval. While almost identical to simple random sampling, systematic sampling can be compromised if the sampling frame is not randomized.

Radey (2020) explains the stratified sampling design as grouping the homogenous subsets of the population before sampling. Stratified sampling increases the representativeness in a sample by using large sample size and similar (homogenous) groups. This design ensures that appropriate numbers of elements are drawn from each subset to reflect representativeness adequately. Standard stratifying variables include age, race, income, and sex.

Cluster sampling is chosen for sampling from particular groups of people or sampling units (Radey, 2020). Groups are sampled first, and members of the group selected for sampling at the next stage.

4)List the advantages and disadvantages of the basic types of probability sampling.

Simple Random Sampling: According to Rubin & Babbie (2017, p. 366), there is no real advantage to using simple random sampling except when a specific, small sampling frame is available. The disadvantages include that it is not the most accurate method available and can be time-consuming.

Systematic Sampling: Systematic sampling is more accurate than simple random sampling and simpler to use (Rubin & Babbie, 2017, p. 366). The major disadvantage is the danger that systematic sampling can be affected by periodicity, the arrangement of the elements arranged in a cyclical pattern that coincides with the sampling interval, and biases the sample (Rubin & Babbie, 2017, p.367).

Stratified Sampling: The advantage of stratified sampling is that it allows for a higher degree of representativeness in a sample and decreases the probable sampling error (Rubin & Babbie, 2017, p. 368). The disadvantage is that it can be tricky to choose variables that are related to variables that the researcher wants to be represented accurately (Rubin & Babbie, 2017, p. 368).

Cluster Sampling: The advantages of cluster sampling are that it is used when it is impossible or impractical to obtain a list of elements of a target population and is highly efficient (Rubin & Babbie, 2017, pp. 371-372). The disadvantages of cluster sampling are that it is less accurate due to sampling error present at each stage of clustering; care needs to be taken to include many homogenous clusters (Radey, 2020).

5)Describe the meaning of sampling error.

Sampling error is defined by Rubin & Babbie (2017, p. 363), as the difference between the true population parameter (a parameter is the summary description of the given variable in a population) and the researcher’s estimate.

6)Explain how samples are used to describe population.

Sampling frames are lists of the elements of a population from which the sample is being selected (Rubin & Babbie, 2017, p. 360). Samples drawn from these frames describe the population that composes the frame and cannot be generalized past that sampling frame. Sampling frames often define the study population, rather than the population setting the sampling frame (p.361).

7)Explain why some randomly selected samples can be biased in light of sampling frame issues and nonresponse.

Sample frames can be biased as they may not always include all the units they are believed to include. Nonresponse bias can occur when participants chose not to participate for various reasons and thus influences the ability to obtain a probability sampling (Rubin & Babbie, 2017, pp. 361-362).

Reliance on Available Subjects: The advantage of this type of nonprobability sampling is that it is easy, accessible, and inexpensive (Rubin & Babbie, 2017, p. 353). The disadvantage is that it is an extremely unreliable and risky sampling method in that it does not account for representativeness and generalizations.

Purposive and Judgmental Sampling: The advantages of this method of sampling are that it allows for the selection of atypical cases, rather than typical cases, where the researcher can use knowledge of the community to handpick participants (Rubin & Babbie, 2017, p. 355). The disadvantage is that, due to a potential lack of meaningful population, this sampling may be better suited for pretest situations rather than full studies.

Quota Sampling: The advantage of this method is that it addresses representativeness (Rubin & Babbie, 2017, p. 356). There are several disadvantages which include difficulty in obtaining accurate quota frames and bias that may exist in the selection of sample elements within a given cell.

Snowball Sampling: This method of sampling is advantageous when members of a target population are difficult to locate (Rubin & Babbie, 2017, p. 356). The disadvantage is that is results in samples with questionable representativeness.

References

Radey, M. (2020). Sampling [PowerPoint Slides]. Canvas@FSU. https://canvas.fsu.edu/courses/128586/files/8237967?module_item_id=1983264

Rubin, A., & Babbie, E. R. (2017). Research methods for social work (9th ed.). Boston: Cengage Learning

Response Post 2:

1)Differentiate between probability and non-probability sampling.

The main difference between probability and non-probability sampling is randomization. In probability sampling, subjects are randomly selected and therefore there is an equal chance that subjects are chosen. A non-probability sample depends on the researcher’s ability to select elements for a sample from available subjects. This type of sampling is used when randomization is unethical or difficult to use. (Singh, 2018).

2)Discuss the logic of probability sampling.

Probability sampling is necessary because of the variations of populations we study. Probability sampling would not be necessary if the population’s elements were all the same. The logic of probability sampling involves techniques to insure the sample adequately represents all of the variations of the total population. (Rubin & Babbie, 2017)

3)Give descriptions of the basic types of probability sampling – simple random, systematic, stratified, and cluster.

According to our text, Simple random sampling involves a number assignment to each element of a sample and the number are selected at random. This sampling technique is used when there little is known about the population (Rubin & Babbie, pp.366, 2017). For example, a church congregation of 1,000 members, in a sample selection of 100, each member has an equal chance of being selected, like pulling names out of a hat.

Systematic sampling involves randomly choosing a subject in intervals. Using the example of the 1000 church members, each member would be an element and would have a number. Systematic sampling would involve choosing every 10th member, making sure to choose the elements at random so that member 10, 20, 30, 40, etc. aren’t chosen. Systematic sampling is easier to administer than simple random sampling. (Rubin and Babbie, 2017)

Stratified Sampling is a modification of systematic and random sampling, not an alternative.(Rubin & Babbie, pp.368, 2017). In this sampling method, prior knowledge of the population is required and the researcher divides the population into groups, and the samples are taken from each group for the sample. In the church example, the researcher may divide the church by different ministries: youth minister, sick and shut in, usher ministry, singles ministry, prison ministry and outreach ministry. Next, individuals from each group are chosen for the sample.

Cluster sampling is similar to stratified sampling as the population is divided into groups. However, in a cluster sample the elements represent the elements in the total population, whereas in a stratified sample the elements in each groups are homogenous. Cluster sampling is more convenient and practical but it is important that each cluster have representative elements to prevent bias or non-representativeness. (Sampling YouTube video)

4)List the advantages and disadvantages of the basic types of probability sampling.

 Type of Sample Advantage Disadvantage Simple Random Easily implemented Ideal Method Produces unbiased samples Expensive Difficult to use when researching human beings Systematic Easier to administer than simple random selection Certain elements can be choses more than others if there is an uneven distribution of elements in the population, creating bias Stratified Strata/Groups Offer good random sampling Difficult to administer Information about the population is needed Cluster More convenient and practical then simple random sampling If the elements within the cluster are not aligned with the elements being studied, bias or non-representativeness occurs Less accurate sample because sampling error is introduced at each stage of the clustering

5)Describe the meaning of sampling error.

The difference between the true population parameter and the researcher’s estimate = sampling error. The larger the sample size, the smaller the likelihood of sample error.

6)Explain how samples are used to describe population.

Samples are used to make inferences about a population. Samples are used in experiments to describe populations in an exploratory, descriptive, or defining manner. A sample is used as a representation of a larger population.

7)Explain why some randomly selected samples can be biased in light of sampling frame issue and nonresponse.

All elements must have equal representation in a sampling frame” (Radney, 2020). Overgeneralization occurs when the sampling frame does not contain all of the elements of the population, yet inference of the population are made based on the small sampling frame. Nonresponse bias occurs when randomly selected subjects chose not to participate. Their responses, or lack of observations of these subjects can skew the results, thus creating nonresponse bias.

 Sample Advantages Disadvantages Accidental/Convenience Sampling Less expensive, more feasible in social work Can provide useful, tentative findings Risky Underrepresentation can occur Judgement/Purposeful Sampling Saves time and money Handpick members of sample group such as experts, to study a population Helps to form a hypothesis or test the effectiveness of questionnaires in an experiment Probability theory can’t be applied Cant be sure if sample is representative when using third party information from experts interviewed Selection is biased Quota Sampling Representative of the population Difficult to get up-to-date information for sample frames Selection bias may occur Snowball Sampling Useful when members of a population are difficult to locate Snowball occurs when subject refers additional subjects Useful for minority and oppressed populations Can’t be sure sample is representative Can’t determine sampling error

References

Nic. (2012, March 13). Sampling: Simple Random, Convenience, systematic, cluster, stratified-Statistics Help. [Video].YouTube. https://www.youtube.com/watch?v=y3A0lUkpAko&feature=emb_rel_end

Radney, M. (2020, July). Sampling [Powerpoint Slides]. https://canvas.fsu.edu/courses/128586/files/8237967?module_item_id=1983264

Rubin, A. & Babbie, E.R. (2017). Research Methods for Social Work. Cengage Learning.

Singh, S. (2018). Sampling Techniques. Medium. https://towardsdatascience.com/sampling-techniques-a4e34111d808