RESEARCH METHODOLOGY
4.7 Sampling procedure
i. Face validity
Govender (2010:163) views face validity as the “realistic outlook of the instrument and it makes the results credible to the common audience”. This was considered when designing the questionnaire.
ii. Construct validity
Construct validity refers to establishing whether a scale or test measures the construct adequately (Govender, 2010:163). In constructing the research instrument the theoretical framework informed the questions asked.
iii. Content validity
According to Govender (2010:163) “content validity refers to the extent to which the measure assesses the broad characteristics of the study”. The questionnaire was valid in content as it measured learning organisation, resistance to change and knowledge on the NEC, which were the constructs under study.
iv. Criterion validity
To measure criterion validity, researchers must calibrate it against a known standard.
This is subject to pre-test and it is used as a standard for judgement. The measuring instrument chosen for this study was subject to pre-test and the resultant adjustments were made.
functional teams responsible for purchasing goods and services in the various operational departments of TNPA.
The following departments were identified and the total number of employees is shown in brackets.
i. Planning, Infrastructure and Engineering (140 employees);
ii. Marine Engineering (55 employees);
iii. Finance (58 employees);
iv. Legal and Compliance (35 employees);
v. Lighthouses Technical (20 employees);
vi. Workshop 24 Technical (25 employees);
vii. Dredging Services ,Engineering (20 employees);
viii. Properties (7 employees);
ix. Operations department (10 employees) and x. Procurement (30 employees)
Thus the target population came to 400.
4.7.2 Access to sample
To access the sample for this study the following procedure was undertaken:
i. Permission to conduct a survey in the organisation was requested from TNPA and was granted by the Training Manager of Transnet National Ports Authority, Durban. (Annexure 3)
ii. A list of target employees was requested and obtained from all Business Unit Managers. The lists were in order of superiority, starting with Senior Managers, Middle Managers, Supervisors and Junior Employees.
4.7.3 Sampling technique
Saunders et al. (2009) suggest two types of sampling techniques, these being probability or representative sampling and non-probability or judgmental sampling.
Probability sampling is associated with survey and experimental research. With probability sampling the probability of each case being selected from the population is equal. Saunders et al. (2009) suggest stages to be followed in the process of probability
sampling which involve the decision on the sample size of the study, representation of the population, selection of the appropriate sampling technique as well as the identification of the suitable sampling frame in line with the research question or the objectives of the study.
There are various kinds of probability sampling that Saunders et al. (2009) have identified, which are:
i. Simple random sampling
This is used when the researcher has an accurate and easily accessible sampling frame where the entire population is listed. Every selected person or element has an equal chance being selected in a sample.
ii. Systematic sampling
This involves the selecting of the sample at regular intervals form the sampling frame, for example the sample will consist of every fourth person and the list is called a sampling frame.
iii. Stratified sampling
The researcher compares different populations in each group chosen for the sample. It is an accurate and easily accessible technique.
iv. Cluster sampling
This refers to the technique of focusing on physical areas within different geographical areas. The focus is on clusters not on individuals.
Non-probability, on the other hand, is based on the assumption that the sample will be chosen statistically at random (Saunders et al., 2009), thus making it to be subjective.
Various kinds of non-probability sampling as suggested by Saunders et al. (2009) are discussed below:
i. Convenience or haphazard sampling
Here the researcher selects people who are the easiest to access for the sample. This technique is convenient and widely used.
ii. Judgemental or purposive sampling
Judgemental sampling occurs when the researcher uses his own judgement in choosing who to approach. This sampling technique is normally used when working with small samples.
iii. Quota sampling
This type of sampling is normally used for a large population to make sure that certain subgroups of units are represented in the sample in approximately the same proportions as they are represented in the population.
iv. Snowball sampling
Snowball sampling is used when it is not easy to identify members of the desired population.
v. Self- selection sampling
According to Saunders et al. (2009:241) self-selection sampling “occurs when you allow each case, usually individuals, to identify their desire to take part in the research”.
For this research the chosen technique was non-probability sampling, in particular, convenience sampling. Sekaran and Bougie (2009) describe convenience sampling as the technique for collecting information from the population group who are conveniently available to provide it. The group of employees from all the operational departments of TNPA, who always take part and form the cross-functional team when goods and services are purchased, was approached for this study.
4.7.4 Sample size
Of the 400 questionnaires that were distributed only 200 were completed and returned and this made the sample size of this study. The response rate of 50% was deemed sufficient for the purpose of this study.
4.7.5 Characteristics of the sample The sample was made up of:
i. Male and female employees.
ii. Employees who were permanently employed at TNPA.
iii. Employees who were 22 years old and above.
iv. Senior Managers, Middle Managers, Supervisors and Junior staff.