Qualitative Studies - Essay Example

Qualitative Studies tend to recruit small samples. Discuss the type of sampling used in qualitative research, and explain the advantages and disadvantages of this kind of sampling. Introduction In general, most qualitative research has an aim of purposive or purposeful sampling; that is, explicitly selecting samples, who it is intended will generate appropriate data. It has been suggested that the overall aim of purposive, as opposed to probability, sampling is to include information-rich cases for in-depth study. To achieve this, a number of different sampling strategies are possible. In qualitative studies, the task of sampling typically takes place iteratively. Usually a qualitative research begins by talking with or observing a small number of people who are known to have first-hand experience with the study phenomenon. Therefore, the discussions and observations are structured loosely, thus allowing for the expression of a "full range of beliefs, feelings, and behaviours" (Polit and Beck, 2007, 57). However, in contrast to quantitative studies, analysis and interpretation are ongoing and concurrent activities that will continue to guide choices about the next sample with concurrent modifications of the types of questions or necessary observations. Thus with the goal of clustering together of related types of narrative information into a coherent scheme of time consuming and intensive data analysis, the sample needs to be determined. Types of Sampling The aim of most qualitative studies is to discover meaning and to uncover multiple realities, and so generalizability is not a guiding criterion as in quantitative studies. Thus probability sampling is not necessary. Randomly selected population may not necessarily be good informants, and thus the sampling method must be one of nonprobability methods. These include: extreme or deviant case sampling, typical case sampling, and snowball sampling (Cochran, 1977, 13-47). This means, there must be a purpose for any particular sampling method. In probability sampling, elements are selected by nonrandom methods. In this method, there is hardly any way to predict the probability that each element is included in the design, and every element does not have a chance of inclusion. There are three primary methods of nonprobability sampling. These are convenience, quota, and purposive (Polit and Beck, 2007, 292). Convenience Sampling A convenience sample can be defined as a sample in which research subjects are recruited based on their ease of availability, or in other words, the sample comprises of the most conveniently available subjects. Essentially, individuals who are the most ready, willing, and able to participate in the study are the ones who are selected to participate. In qualitative research, it may be helpful to use a convenience sample to test the appropriateness of interview questions. This is an inexpensive and quick way to test the design of the study by approaching an interested group of people first before embarking on a larger, longer, and more expensive study (Diekmann & Smith, 1989, 418-430). The main problem with such sampling is that the available subjects might be atypical of the population of interest with regard to critical variables. It has been argued that convenience samples can lack transferability or external validity in qualitative research. Thus although this is simple and more cost-effective, it is important to remember that the participants recruited are not necessarily reflective of the population being studied, and they may not necessarily be reflective of all view points, and thus the research may be criticized to be biased (Polit and Beck, 2007, 292). Quota Sampling (Stratified Purposeful) This is also known as quota sampling in which the researcher identifies population strata and also determines the number of participants needed from each stratum. To be able to determine this, the information about the population characteristics is necessary, since it should represent diverse segments preferable in a representative proportion of occurrence in the population. This sampling is a nonprobability method since it requires only that the quota for each category be met without any further attention to how those sample members are actually located. Quota samples are not truly generalizable because even though the quota-based categories may match their size in the larger population, the sample may be unrepresentative on other characteristics outside the quota system (Levey & Lemeshow, 1980, 16-23). In qualitative research, quota sampling is a quicker and cheaper alternative to random sampling. This specific method may be used to define the data categories that have eligibility for study. This method is often connected to stratified sampling as a specific approach to purposive sampling (Polit and Beck, 2007, 293). Purposive Sampling The way the researchers sample must be tied to their purpose and the objectives of the research. This also indicates there cannot be a best strategy of sampling because the sampling method will link to the context and nature of the research objectives. Purposive sampling is virtually synonymous with qualitative research. However, because there are many objectives that qualitative researchers might have, the list of purposive strategies that may be followed is virtually endless, and any given list will reflect only the range of situations the author of that list has considered (Mercorio, 2003, 1291-1293). Literally, the researcher is able to hand pick recruits judged to be typical of the population or specifically knowledgeable for the study. This is based on the researcher's knowledge on the topic, and therefore, the sampling is essentially subjective, but this method does not provide any external or objective method of assessing how the subjects become typical of the study (Polit and Beck, 2007, 294). Maximum Variation Sampling Searching for cases or individuals who cover the spectrum of positions and perspectives in relation to the phenomenon the researcher is studying and would include both of the extreme and typical cases plus any other positions that can be identified. This means this involves a purposive design that includes a wide range of variation on different dimensions of interest. Although it is possible that selection of subjects with wide variations in interests and views would challenge the thematic conceptualization, but this can lead to an array of viewpoints about the phenomenon studied (Polit and Beck, 2007, 306). Homogeneous Sampling This sampling methodology intends to sample in such a manner that it deliberately reduces variation. Reduction of variation would create provisions for a more focused enquiry. When a particular characteristic of a population is studied in detail and in depth, this technique is used (Moody et al., 1988, 374-379). This is often used to select people for group interviews. In this way the research process becomes simplified, and the magnitude of the research process is reduced (Polit and Beck, 2007, 306). Critical Case Sampling In this sampling method, this facilitates the researcher who might be looking for a decisive case that would help make a decision about which of several different explanations is most plausible, or is one that is identified by experts as being a particularly useful choice because of the generalizations it allows (Morse, 2000, 3-5). This means the researcher selects the important cases in relation to the phenomenon of interest. A particularly suitable case that highlights the critical aspects of the phenomenon is chosen (Polit and Beck, 2007, 307). Theory Based Sampling Researches which follow a more deductive or theory-testing approach would need samples of individuals or cases embodying theoretical constructs of the question. This actually is a particular type of criterion sampling, but there may be overlaps between these categories (Sandelowski, 1995, 179-183). This means individuals or cases are selected in the sampling which would potentially represent these theoretical constructs. Therefore, it would need a focused approach to examine a theory from the qualitative angles (Polit and Beck, 2007, 307). Confirming and Disconfirming Cases Sampling This tests the theory and failure to prove it would confirm it. Towards the end of the data collection in qualitative studies, from the trends or patterns of data, the conceptual format may be checked on the basis of such sampling. Confirming cases would support additionally the conceptual framework of the researcher and would enhance credibility. With disconfirming cases sampling, the researcher employs a strategy that looks forwards to extend the analysis by looking for cases that will disconfirm it (Polit and Beck, 2007, 307). Snowball or Chain Sampling Snowball sampling uses an initial set of data sources as the basis for locating additional data sources. Therefore an initial pool of small informants is used who nominate the other recruits meeting the inclusion criteria of the study. This is a method of nonprobability sampling. This is a useful way to pursue the goals of purposive sampling in many situations where no lists of potential participants are available (Vervaeke et al., 2007, 1705-1713). Typically, this process starts with interviewing an initial set of participants who serve as informants of the research topic and also about other potential participants. There is a distinct risk of capturing a biased subset of the total population (Polit and Beck, 2007, 292). Extreme or Deviant Case Sampling This design provides opportunity to learn from the unusual and most extreme informants, such as informants with notable failures and extreme successes. This method of sampling may be successful since there is usually a rich collection of information from these special informants. Although there is risk of distortion of the context of the phenomenon, observation of these samples usually generates a wealth of information (Polit and Beck, 2007, 306). Typical Case Sampling Sampling methods involving the average or typical participants usually generate a typical qualitative profile that illustrates the normal, usual, and typical manifestations of the phenomenon under study (Polit and Beck, 2007, 307). Intensity Sampling This sampling methodology is similar to extreme case sampling but usually do not look into either of the extremes. This means samples from information rich cases are drawn which tend to intensely express the phenomenon of interest. Thus it takes cases other than those where information may be potentially distorted, but rich enough to offer information that strongly illuminated the phenomenon (Polit and Beck, 2007, 307). Politically Important Cases Another interesting strategy suggested is 'political sampling', or taking into account the political considerations that apply to both sample size and selection (Porter, 1999, 796-804). It has been suggests that choosing politically important cases is one strategy for improving the chances of a project gaining attention and the findings being used (Polit and Beck, 2007, 307). Random Purposeful Sampling When purposive sampling is done in a random manner, this is called random purposeful sampling. Randomization involves a process in which all subjects in the population have an equal and independent chance of being selected (Brown et al., 2008, 178-181). When the sample size is large in a qualitative design, the randomization leads to credibility of the research (Polit and Beck, 2007, 297). Stratified Purposeful Sampling In this sampling method, the population is first divided into two or more strata. This enhances representativeness of the sample through homogeneous sub-division of the population into subsets from which appropriate number of samples is selected randomly (Sandelowski, 2000, 246-255). This is done by grouping together elements that belong to a particular stratum and having a random selection of the desired number of elements (Polit and Beck, 2007, 297). Criterion Sampling This involves searching for cases or individuals who meet a certain criterion. Thus studying cases with the predetermined criterion provides insight to the phenomenon. These are used sometimes used in mixed method studies (Polit and Beck, 2007, 307). Opportunistic Sampling This is a sampling method where within the study, the researcher remains open to a new lead about the phenomenon as it unveils (Russell and Gregory, 2003, 36). Purposefully, then the researcher takes the opportunity of this new fact and includes new samples that fit to the new criteria (Polit and Beck, 2007, 306). Combination or Mixed Sampling When the sampling method in a particular study uses two or more basic methods of sampling, it is called mixed sampling. These naturally meet multiple interests and needs. Moreover, mixing methods lead to flexibility to the design, where the phenomenon may be studied through triangulation (Teddlie and Yu, 2007, 77-100). Heterogeneous Sampling A heterogeneous sample contains individuals or groups who differ from each other in a major aspect. This is also known as maximum variation sampling. This involves searching for cases or individuals who cover the spectrum of positions and perspectives in relation to the phenomenon being studied and would include both the extreme and typical cases plus any other positions that can be identified (MacDougall & Fudge, 2001, 117-126). This invites challenges to preconceived or emerging conceptualizations (Polit and Beck, 2007, 306). Conclusion This discussion indicates that there are a large variety of sampling techniques in qualitative research. Some of these have been just outlined here. Depending on the phenomenon studied, the sampling technique must be adjusted and selected; however, knowledge about the phenomenon and a clear understanding about the sampling methods will deliberate the most accurate results which will have validity and generalizability. Reference Brown, C, Hofer, T, Johal, A., Thomson, R., Nicholl, J., Franklin, BD., and Lilford, RJ, (2008). An epistemology of patient safety research: a framework for study design and interpretation. Part 4. One size does not fit all. Qual. Saf. Health Care; 17: 178 - 181. Cochran, W. G. (1977). Sampling techniques (3rd ed.). New York: John Wiley & Sons. 13-47. Diekmann, J. M., & Smith, J. M. (1989). Strategies for accessment and recruitment of subjects for nursing research. Western Journal of Nursing Research, 11, 418-430. Levey, P. S., & Lemeshow, S. (1980). Sampling for health professionals. New York: Lifetime Learning. 16-23. MacDougall, C., & Fudge, E. (2001). Planning and recruiting the sample for focus groups and in-depth interviews. Qualitative Health Research, 11, 117-126. Mercorio, F., (2003). Differing results for efficacy--sampling bias or purposive sampling Fertil Steril; 80(5): 1291-2; author reply 1292-3. Moody, L. E., Wilson, M. E., Smyth, K., Schwartz, R., Tittle, M., & VanCott, M. L. (1988). Analysis of a decade of nursing practice research: 1977-1986. Nursing Research, 37, 374-379. Morse, J. M. (2000). Determining sample size. Qualitative Health Research, 10, 3-5. Polit, DF and Beck, CT., (2007). Nursing Research Principles and Methods. 7th Edn. Lippincott and Williams and Wilkins. London. pp. 291-308 Porter, E. J. (1999). Defining the eligible, accessible population for a phenomenological study. Western Journal of Nursing Research, 21, 796-804. Russell, CK and Gregory, DM., (2003). Evaluation of qualitative research studies. Evid. Based Nurs.; 6: 36. Sandelowski, M. (1995). Sample size in qualitative research. Research in Nursing & Health, 18, 179-183. Sandelowski, M., (2000). Combining qualitative and quantitative sampling, data collection, and analysis techniques in mixed-method studies. Res Nurs Health; 23(3): 246-55. Teddlie, C. and Yu, F., (2007). Mixed Methods Sampling: A Typology With Examples. Journal of Mixed Methods Research; 1: 77 - 100. Vervaeke, HK., Korf, DJ., Benschop, A., and van den Brink, W., (2007). How to find future ecstasy-users: targeted and snowball sampling in an ethically sensitive context. Addict Behav; 32(8): 1705-13. Read More