Goals

After reading this article, you will be able to do the following:

• Identify the different sampling method types.
• Define sampling bias.
• Explain the difference between convenience sampling and purposive sampling.

Introduction

This article focuses on why you must pay attention to sampling issues when developing your research. You have decided what questions you want to answer in your study. You have read articles or talked to professionals, getting vital information to guide your research. It is time to choose your study participants. Participants are selected from the group that the collected data is intended to describe. This group is called your research population (Lippman, 2017). However, because it is complicated, time-consuming, and impractical to include the total population in the study, you need to get a members’ proportion from the targeted group (Thompson, 1999). This population subset is called a sample. A sample is a smaller subset of the entire population, ideally one that reasonably represents the whole population (Lippman, 2017).

When completing any research, you must choose your sample carefully to minimize bias. The term bias describes statistics that do not accurately represent the population, causing distorted results and wrong conclusions (Simkus, 2023). You must consider three essential study components to determine your sample. First, the question types that you ask for conducting your research will determine the more appropriate sampling method to use to choose your sample. Second, determine the research type that you are conducting—quantitative or qualitative.

Qualitative research is used to gain insights into phenomena, groups, or experiences that cannot be objectively measured or quantified using mathematics. It is more exploratory research, such as ethnographic studies, focus groups, and recollections of events or experiences.

Quantitative research provides precise causal explanations that can be measured mathematically. It is conducted in a controlled environment to obtain objective information, such as polls, surveys, and experiments (National University, n.d.). You will learn what sampling methods are better for each study type.

Third, the participant number needed depends on your research type and sampling method. Thompson (1999) explained that, although there is no strict rule to choosing the participant number, for quantitative research, at least thirty participants are recommended. This number is based on the law of small numbers, which stipulates that thirty participants allow adequate hypothesis testing. However, if the population is small, use as many participants close to the population as possible. For qualitative research, the participant number could be any number depending on the study scope since it is exploring subjective experiences. This article focuses on why you must pay attention to sampling issues when developing your research.

Random Sampling Methods and Sampling Bias

Imagine researching students’ online course experience, however, your sample includes students who have never taken an online course. Inevitably, the results leave out essential research data, leading to unreliable outcomes and misleading conclusions. Therefore, after all your work is done, your findings cannot be used to demonstrate or represent the sample populations’ ideas and thoughts (Simkus, 2023). There are many ways to sample a population, but there is one goal to keep in mind: the sample must be representative of the population (Lippman, 2017).

One way to ensure that the sample reasonably mirrors the population is to employ randomness. The most basic random method is simple random sampling. A simple random sample is one in which every population member and any member group has an equal probability of being chosen (Lippman, 2017).

An example of selecting the online course-experience study sample is to put all the students’ names who registered for online courses in a bag and draw 200 names without putting them back in the bag again. Of course, this will take quite some time, yet it will give everyone the same chance of being chosen. There are other ways that you can do this without so much work. Salt Lake Community College has a Division of Institutional Effectiveness office slcc.edu/institutional-effectiveness/dsa/index.aspx that can help you choose your sample via a computer.

Sampling bias is when the sample is representative of something other than the target population that you want to study (Lippman, 2017). In the students’ online course experience study, a sampling bias results when your sample includes students who have never taken an online course. Two main reasons cause the sampling bias: poor methodology and poor execution.

Poor methodology happens when the researcher does not set clear guidelines for the intended study population. For example, for the student online course experience study, it is essential to determine how to choose the participants representing the targeted group. Also, it is best to consider other variables, such as including all online students or only those registered for a specific semester. Remember that through random sampling, you can prevent sampling bias when recruiting participants because the target population members will have the same opportunity to be chosen. Poor execution happens when the researcher uses a very small sample size or gives up on reaching non-responders, jeopardizing the methodology’s setup.

When conducting quantitative research, the best way to prevent sampling bias is to use simple random sampling (discussed previously), stratified sampling, cluster sampling, and systematic sampling because every member has an equal probability of being chosen (MasterClass, 2022). Let’s discuss these methods in more detail.

Stratified sampling divides the intended population into several subgroups or strata. Random samples are then taken from each subgroup with sample sizes proportional to the population’s subgroup size. Stratums are formed based on members’ shared, unique characteristics such as age, income, race, or education level (Lippman, 2017). For the student online course experience study, you decide to include all students taking online courses at SLCC. So, the online students are divided into first-year, second-year, third-year, and returning students. Then, 25 percent of each group is selected to be final sample participants. It is essential that from the 25 percent selected, you consider having your population exactly represent gender, ethnicity, social background, and other variables that can affect the results. This is this method’s main disadvantage.

Cluster sampling is a method where researchers divide a large population into smaller groups, known as clusters, and then select randomly among the clusters to form a sample.

Cluster sampling is typically used when the population and the desired sample size are particularly large. For example, a researcher divides a city into ten school districts and chooses three or four whole school districts as the sample. Remember, stratified sampling divides the population into subgroups (strata) based on shared, unique member characteristics. Then, members are randomly selected from the subgroups to form the sample. Cluster sampling will divide the population into subgroups or clusters, such as city blocks or school districts, and then randomly select a few whole clusters to be a part of the sample (Simkus, 2023).

Systematic sampling selects every nth population member to be in the sample. The nth member means the person’s position in a sequence. For example, you will choose every fifth person in a fast-food line to survey. Or, you select a sample using systematic sampling—a pollster uses a phone book and chooses every one-hundredth name to call (Lippman, 2017). It would be best if you were careful about how the population is listed. For example, your population list has men with even numbers and women with odd numbers. You choose to sample every tenth individual. Therefore, only men will be included in your sample (Thomas, 2020). Another example of a sampling bias with this method is when you want to ask one hundred customers what their favorite breakfast item is. You go to a grocery store and ask the first one hundred customers who enter the place. The problem is that you need to consider the customers who shop in the afternoon or evening (Wells & Payne, 2023). Therefore, your study does not consider their answers, and you need to consider a better population sample that shops in the selected grocery store.

Non-Random Sampling Methods and Sampling Bias

Let’s now discuss four sampling types that are not selected randomly such as the ones discussed previously. These sampling types are used mostly for qualitative research. They are convenience sampling, purposive sampling, voluntary sampling, and snowball sampling. In convenience sampling, the sample is chosen by selecting whoever is convenient. The participants are recruited primarily because they are available, willing, or easy to access or contact (Lippman, 2017). This sampling type is used when no criteria are required to be part of this group. For example, a university student working on a project wants to understand the favorite soda on campus on a Friday night and will mostly call their classmates and friends and ask what soda they consume.

According to the Dovetail Editorial Team (2023), purposive sampling is often used in qualitative research to select a specific group with the characteristics or attributes the researcher is interested in studying. The participants are chosen not randomly. This is why this method is also known as selective sampling. For example, you want to study people’s experiences with chronic pain. You might use purposive sampling to select a sample of individuals diagnosed with chronic pain.

What is the main difference between convenience and purposive sampling? In purposive sampling, the participants are selected based on defined attributes that the researcher is interested in. In convenience sampling, the sample chosen is readily available or easily accessible to the researcher, not based on any specific characteristics. Both methods can introduce a sampling bias because the researcher may unconsciously select individuals who fit their expectations, affecting the study’s validity (Dovetail Editorial Team, 2023).

Voluntary sampling is the method where the study participants volunteer. The researcher can seek the sample by person, internet, public posting, and other methods. For example, a radio show host asks the listeners to go to their website and complete a poll. This method’s advantage is that it ensures that only those interested in the study will participate, and it is cost-effective. On the other hand, since the sample is not random, voluntary sampling can cause biased results that do not represent the population. Also, it can lead to response bias because participants can give answers that they think the researcher expects (Williams, 2023). For example, a TV show host asks the viewers to visit the show’s website and respond to an online poll.

The snowball-sampling recruitment technique asks the research participants to help the researcher identify other potential participants. It would be best if you justified using this method according to the study context and target population. For example, your study requires responses from previously incarcerated people. Finding an adequate number of participants and recruiting them yourself could be challenging. However, if you get a few previously incarcerated people to participate in your study, they could help you to get in touch or recruit other previously incarcerated people.

Among some snowball-sampling advantages are that you can easily find suitable participants for your investigation using the referral system, and it allows you to collect responses from people who would have hesitated to participate in your research. Nevertheless, this sampling method can cause sampling bias due to the interest population’s lack of representation since you rely mainly on referrals.

Conclusion

You have learned what sampling method types you can use for quantitative or qualitative studies. Also, you have the knowledge to prevent sampling bias depending on the sampling method you want to use. Now, with the research type you want to complete and the questions you want to answer, it is time to gather your study participants. You’ve got this!

Practice

The following questions will help you to review part of the information we have discussed.

1) Determine which sampling method is used in the following situations:

1. a) Every fourth person in the class was selected.
2. b) A sample was selected to contain twenty-five men and thirty-five women.
3. c) Viewers of a new show are asked to vote on the show’s website.
4. d) To survey voters in a town, a polling company randomly selects ten city blocks and interviews everyone who lives on those blocks.
5. e) A sample is chosen by pulling fifty names from a hat.

2) What is sampling bias?

3) What is the main difference between convenience sampling and purposive sampling?

References

Dovetail Editorial Team. (February, 2023). What is purposive sampling? Retrieved from https://dovetail.com/research/purposive-sampling/

Lippman, D. (2017). Math in Society (Ed. 2.5). Pierce College, Ft Steilacoom.

MasterClass. (2022, Feb 24). 6 types of sampling bias: How to avoid sampling bias. Retrieved from www.masterclass.com/articles/sampling-bias

National University. (n.d.). What is Qualitative vs. Quantitative Study? Retrieved from https://www.nu.edu/blog/qualitative-vs-quantitative-study/#:~:text=While%20both%20share%20the%20primary,how%20different%20people%20experience%20grief.

Oregon State University. (2010). Snow sampling. Retrieve from https://research.oregonstate.edu/irb/policies-and-guidance-investigators/guidance/snowball-sampling#:~:text=Snowball%20sampling%20is%20a%20recruitment,in%20identifying%20other%20potential%20subjects.

QuestionPro. (n.d.). Convenience sampling: Definition, advantages, and examples. Retrieved

from https://www.questionpro.com/blog/convenience-sampling/#:~:text=Researchers%20use%20convenience%20sampling%20in,a%20part%20of%20this%20sample.

Simkus, J. (2023, July 31). Cluster sampling: Definition, method, and examples. Retrieved from https://www.simplypsychology.org/cluster-sampling.html

Simkus, J. (2023, July 31). Sampling bias: Types, examples & how to avoid it. Retrieved from https://www.simplypsychology.org/sampling-bias-types-examples-how-to-avoid-it.html

Thomas, L. (2020, October 2). Systematic sampling: A step-by-step guide with examples. Retrieved from https://www.scribbr.com/methodology/systematic-sampling/

Thompson, C. (1999). If you could just provide me with a sample: examining sampling in qualitative and quantitative research papers. Evidence-Based Nursing (Vol 2, No 3). 68 – 70. doi: 10.1136/ebn.2.3.68.

Wells, D., & Payne, T. (2023). Sampling bias definition, types & examples. Retrieve from https://study.com/learn/lesson/sampling-bias-examples-types.html

Williams, Kate. (November, 2023). Volunteer sampling: Insights, applications, advantages. Retrieved from https://surveysparrow.com/blog/volunteer-sampling/

Answer key

1) a. systematic, b. stratified, c. voluntary, d. cluster, e. simple random.

2) The sample is not representative of the population.

3) In convenience sampling, the sample is selected by accessibility, in purposive sampling, the sample is selected by defined attributes.