Page 224 - PhD GT

10.2.1 The study sample
Since the study was concerned with clinical applicability, it was conducted in a clinical population. The sample, in research of this nature, refers to a subset of the population that is used to gain information about the entire population; according to Henry (1990) it is a model of the population. “A good sample will represent the population well” (p.11) and the extent to which it is able to do this will determine the extent to which findings for the sample can be used to generalise to the target population as a whole. The ability of the study to do this is referred to as its external validity. External validity is a function both of the method of sampling and of the way in which the data are obtained from the sample. The method of sampling has implications not just for the generalisability of the findings but also for their statistical validity. The size of the sample determines the ability to conduct tests that demonstrate relationships and changes in the data; for instance when the sample is of insufficient size for the purpose of crossing the threshold of statistical significance, a Type II error may occur: the null hypothesis cannot be rejected when it is in fact false.
Of central concern to the present study were the implications of the nature of change in dependence for treatment planning and the measurement of outcome; the target population was therefore described as those individuals seeking help for problems of substance misuse and dependence. The study population is the sample chosen to represent the target population, and the quality of the sample selected is determined by its similarity with the target population. A number of methods are available for sample selection. In the present study, a convenience sample was chosen as being able adequately to represent the target population of clinic attenders and the one most likely to yield an adequate number of study participants given the available resources. A convenience sample is described by Henry (1990) as a type of non-probability sample in which cases are selected on the basis of their availability for the study. In a probability sample, every member of the population has a possibility of being included in the sample; this is generally not the case in a non-probability sample where subjective judgements may play a role in the selection of the sample.
Threats to the validity of the study findings result from a number of sources of which sampling bias is one (Henry 1990). The clinical sample selected for the purpose of the present study consisted of consecutive attenders at a specialist agency for the treatment of alcohol and heroin use and dependence. The choice of a single site for recruitment to the study was based upon a calculation of the representativeness of such a sample of the total clinical population of problem alcohol and drug users and the combination of available resources with the availability of cases for
219