Clinical research 3: Sample selection☆
Introduction
This paper addresses two key principles of sampling: adequacy (the number of participants) and appropriateness (the characteristics of the participants). The research aims and design will determine whether you need a statistically based or saturation based sample size (adequacy). The appropriateness of the sample is addressed through inclusion and exclusion criteria. These two aspects are discussed below in relation to quantitative and qualitative designs.
Section snippets
Quantitative designs
By definition, quantitative designs rely on the control or exclusion of all factors that might bias the outcome of a study. In the reality of clinical research, this is rarely achievable. However, a crucial early stage is the identification of confounding variables and, for each one, to identify:
- (a)
whether it can be controlled;
- (b)
if not, can it be measured or;
- (c)
should it be used as an exclusion criterion?
Confounding variables that are measurable but not controllable provide useful data for sub-set
Sampling methods
Ensuring that the sample accurately represents the larger population is more important than the size of the sample in quantitative research. The representativeness of the sample in terms of the larger population has a direct impact on the external validity of the conclusions of the research.
The first step in sampling then is to identify the specific characteristics of the target population, i.e. the population to which you wish to generalise the results. For example, a target population may
Calculating sample size
Sample size calculation in quantitative research depends on a number of factors. These include:
- 1.
research design;
- 2.
sampling method;
- 3.
the degree of precision required;
- 4.
the variability of the factors being investigated;
- 5.
the incidence of a particular variable in the population.
As a general statement, the larger the sample the higher the likelihood that the findings will accurately reflect the population because larger samples have lower sampling error. Sampling error refers to the notion of estimating
Controlling for bias
The best way to evaluate the suitability of a sample in terms of its representativeness of the target population is to scrutinise the sampling method and evaluate the size of the sample. Potential biases in sampling can be reduced by rigorous recruitment strategies that increase response rate and reduce the likelihood of systematically excluding some people from the sample or over representing others.
Qualitative designs
Qualitative designs rely on saturation based sample selection to satisfy the sampling principle of adequacy. Appropriateness of sampling is usually achieved through purposive or theoretical sampling.
The concept of saturation
Sampling to the point of saturation requires the researcher to continue to recruit participants until no new data emerge. However, ethics committees require an indication of likely recruitment; five to eight participants are usually sufficient for a homogenous sample and 12–20 for a heterogenous sample, where it is important to maximise variation across the sample (Kuzel, 1999, p. 42). When exploring the information needs of relatives in ICU, you may wish to include relatives of both long and
Purposive sampling
Purposive sampling is commonly used to allow the researcher to include participants with key experience of the issue to be studied. Decisions about whom to sample are typically made before data collection commences. Again, a key decision is whether you need a homogeneous or heterogeneous sample. Also bear in mind that the synergy you would wish to create in a Focus Group interview may require you to sample a multi-professional team.
Specific qualitative methodologies also assist in purposive
Theoretical sampling
With theoretical sampling, the process of data collection is guided by the emerging theory and sampling decisions are taken during data collection.
Controlling for bias
The very nature of qualitative research, with its focus on deliberate targeting of the sample, lends itself to criticisms of bias. Unfortunately, discussion related to the process of identifying saturation is frequently omitted from published research. One approach frequently used to reduce bias is triangulation. The traditional definition of triangulation uses the analogy of navigation, where knowledge of two points is enables calculation of a third. This is often referred to as triangulation
Summary
Regardless of the research approach taken, adequacy and appropriateness of the sample are essential components of research design. It is equally essential that these processes are described for consumers of research, both to de-mystify the steps involved and, more importantly, to allow the reader to judge the credibility of the findings (Selby et al., 1990). All decisions made by the researcher about sample size should be made explicit and the steps taken to control bias should be described.
References (8)
- et al.
The effect of pressure bandaging on complications and comfort in patients undergoing coronary angiography: a multicenter randomized trial
Heart Lung: J. Acute Crit. Care
(1998) Clinical research 1: research questions and design
Intens. Crit. Care Nurs.
(2004)- et al.
Evaluation of sampling methods in research reported in selected clinical nursing journals: implications for nursing practice
J. Prof. Nurs.
(1990) Statistics and ethics in medical research III. How large a sample?
Brit. Med. J
(1980)
Cited by (0)
- ☆
This article was originally published in Intensive and Critical Care Nursing 2005 21(1) 51–55. The article is republished with permission from Intensive and Critical Care Nursing.