There are many sampling strategies that can be used when conducting research. The sampling strategies vary depending on whether the study is qualitative or quantitative. This post discusses the sampling strategies used in qualitative research.
Sampling is the process by which a researcher selects a group of individuals, organisations or units to be included in his study from the target population.
The aim of qualitative research is to create a deeper understanding or meaning of the phenomenon being studied. Therefore the sample selected should reflect this.
Components of the sampling procedure
When developing the sampling procedure, the researcher should include the following four main components:
The people to be studied
These are the individuals who will provide the information required through participation in the study.
Whereas qualitative research is unstructured in nature and most often evolves during the fieldwork, the researcher should have an idea of who his participants will be at the initial stage. This may change as the research progresses.
The time the study will take place
This is an important consideration and varies depending on the aims of the study and the study participants.
Will the study be conducted during weekdays, weekends, school holidays, in the morning, in the evening etc?
The researcher should think critically about this aspect because it will influence the amount of information he will gather from his participants.
Events or processes to be studied
Will the researcher study routine events or processes or will he study special events that occur during specific times?
The study setting
The researcher should consider where the study will be conducted.
For qualitative research, the natural setting of the participants is the most recommended study setting.
For instance, if the study is about the relationship between teachers and students, then the natural setting would be a school.
If the study is about managers-employees relationships, then the natural setting would be the workplace.
The advantage of studying participants in their natural setting is that the researcher can observe what is going on and this would help to validate the responses he gets from the participants.
Sampling in qualitative research
In qualitative research, sampling is less structured and rigid compared to quantitative research. Qualitative research employs purposive sampling techniques because the researcher is interested in participants who will provide rich information that is relevant to what he is studying.
Sampling in qualitative research is also flexible in the sense that if the researcher is in the process of fieldwork and finds that some of the respondents are not providing rich information, he is at liberty to abandon those respondents and recruit new respondents, as long as the reasons are clearly documented.
The sample selected should be adequate; that is, the composition and size of the respondents should be appropriate.
For instance, if your study is about customers’ experience with a particular brand, then your sample should include both customers and managers of the brand. If your study is about employees-managers relationship, then the sample should have both employees and the managers, rather than just employees or managers only.
The appropriateness of the sample helps to judge whether the study findings are of high quality and trustworthy.
In qualitative research, the sample size is normally small compared to the sample size in quantitative research. This is because the data collected from qualitative research tends to be a lot. Additionally, qualitative research focuses on the depth of the phenomenon being studied, rather than its breadth which is the aim of quantitative research.
Nevertheless, the sample size in qualitative research has been the subject of debate. While there is no clear-cut answer to the sample size that qualitative researchers should use, the general agreement is that the sample size determination should be based on information redundancy or saturation; that is, the researcher should stop engaging more participants when no new information is forthcoming.
The researcher should also ensure that the data gathered from the participants is adequate enough to answer the research questions he set out to answer. If carefully selected and with good data collection techniques, a sample size can be able to achieve this.
Some experts advise that individual interviews for qualitative research should not have more than 50 participants because data analysis in qualitative research is quite complex.
In addition to the sample size for individual interviews, the researcher should also consider the number of participants in a focus group discussion (FGD). Ideally, each FGD should not have less than 5 participants and not more than 10 participants.
Common sampling techniques used in qualitative research
All sampling techniques for qualitative research are purposive in nature. That is, samples are selected with the purpose of providing rich information. This is different from quantitative research, where samples are selected with the aim of generalisation of the study findings to the target population.
Sampling techniques used in qualitative research include:
This strategy is the least rigorous because samples are selected because they are the easiest to access. It is therefore convenient to the researcher.
For instance, for a study exploring parents-teenagers relationship, the researcher may choose to select his neighbours with teens because they are closest to him.
The disadvantage with convenience sampling is that the researcher may fail to get information-rich cases.
Theory-based/operational construct/theoretical sampling
In this strategy, the researcher selects individuals or units that manifest an emerging theory from initial data collection and analysis.
Based on the emerging theories and concepts, the researcher will select additional individuals or organisations that manifest them and seek additional information in order to unearth variations of the concepts among different sub-groups.
This sampling technique is mostly used in grounded theory research.
Extreme or deviant case sampling
In this strategy, the researcher will select samples of individuals or units that are extreme cases of the phenomenon under investigation.
For instance, if a study is about factors that influence the success of businesses, the researcher will select samples of business that have collapsed.
This sampling technique selects cases that manifest the phenomenon intensely but not extremely.
The respondents may be average but with rich information hence the researcher is able to learn about the phenomenon intensely.
For instance, if a study is about factors that influence successful scale-up and sustainability of health-tech enterprises, rather than just interviewing the health-tech enterprises only, a researcher may choose to interview organisers of the accelerator program for health-tech companies that helps start-up companies scale their innovations. These organisers would have very rich information about the issue being studied because they are involved in nurturing start-up health-tech companies and enabling them to successfully scale-up their innovations.
Heterogenous/maximum variation sampling
The researcher selects a wide range of individuals/units that are affected by the phenomenon of interest.
This strategy enables the researcher to collect data that describes and explains the key themes and patterns that cut across the diverse groups of individuals or units as well as the unique attributes from them.
For instance, in the study of the factors that influence successful scale-up and sustainability of health-tech companies, the researcher will include in his sample: start-up companies, established companies, locally-owned companies, foreign-owned companies, male-headed companies and female-headed companies.
The researcher selects only cases that are similar in pre-defined characteristics.
The advantage of this technique is that it helps to reduce variations and makes the analysis process easier.
It is mostly used when conducting focus group discussions.
In the health-tech study example, the researcher will select start-up companies only or established companies only etc and collect data from them.
This strategy helps the researcher gain a deeper understanding of the experiences of a particular sub-group e.g. experiences of and challenges facing start-up companies.
Snowball or chain sampling
This strategy is used when it is difficult for the researcher to find the members of the target population.
The researcher will use one or two members to help him identify other members they know of that are information-rich.
The most crucial task for the researcher therefore is to find his entry-point after which the entry point will lead him on to other useful cases.
Critical case sampling
In this strategy, the researcher selects cases that will provide critical information.
The focus of the data collection is to understand what is happening in each case so that generalisation can be made to other cases.
For instance, a study examining vaccine hesitancy amongst a target population will want to first understand hesitancy amongst medical practitioners. If medical practitioners are hesitant towards the vaccine, then it is most likely that other members of the population will also be hesitant.
In this technique, the researcher defines some set of criteria and then selects cases that meet those criteria.
Opportunistic or emergent sampling
This strategy is used in the course of fieldwork when a researcher begins to see some emerging or unexpected themes that he had not anticipated.
The researcher will therefore look for new respondents to provide information about the emergent or unexpected theme.
The strategy depicts the flexible nature of qualitative research.
Typical case sampling
It is the opposite of the extreme case sampling in that the researcher collects data from the average cases rather than extreme cases.
There must be a general agreement on what the average of the cases being studied is. For instance, average household, average income etc.
Stratified purposeful sampling
The target population is stratified by various characteristics and the sample from each strata is selected carefully to ensure that they are relevant to the research question being answered.
For instance, a study on effects of covid-19 on household income, the researcher will stratify the population into low-income, middle-income, and high-income earners, and carefully select cases from each strata to collect the data from.
Purposeful random sampling
In this strategy, the researcher will select a small sample of cases from a larger pool of cases that are relevant to the study.
The cases will be selected at random. Unlike in quantitative research, this strategy used in qualitative research will select a small sample size to add credibility to the study, rather than ensure representativeness of the sample to the population.
For instance, if you want to study the effect of a capacity-building program on the participants, you select a few of the participants at random and collect data from them.
Combination or mixed purposeful sampling
In this strategy, the researcher uses a combination of two or more of the sampling techniques discussed above.
This happens quite often. It is common to use two or more sampling techniques to get a sample that is adequate enough to a study.
In conclusion, qualitative studies employ purposeful sampling techniques with the aim of getting rich information that provide answers to the researcher’s questions.
Whatever sampling technique the researcher chooses, there must be clear justification for choosing that technique and not others.