• Tidak ada hasil yang ditemukan

CHAPTER 3: RESEARCH METHODOLOGY

3.4. Research methodology

51 telecommunications services are not dependable, ix) company has concerns about internet security, x) company has concerns about legal issues, contracts and/ or liability.

52 maintain that in cluster sampling, clusters must be as similar as possible to one another. In addition, clusters must equally enclose a heterogeneous mix of individuals.

v) Systematic sampling: entails selecting individuals according to a predefined sequence (Sekaran and Bougie, 2010).

However, Sekaran and Bougie (2010) argue that when time and other factors become critical (rather than the generalisation of the findings), a non-probability sampling design can be adopted . In non-probability sampling design, there is no guarantee that each segment within the population will be represented. There are also different types of non-probability sampling as there is for probability sampling.

i) Convenience sampling: also known as accidental sampling, involves the collection of information from members of the population “conveniently available to provide it” (Sekaran and Bougie, 2010:276). It is considered to be the best way to collect basic information quickly and efficiently. However, It is considered to be the least representative sampling technique (Leedy and Ormrod, 2010).

ii) Quota sampling: involves selecting respondents in the same proportion that they are found in the population but not randomly (Leedy and Ormrod, 2010).

iii) Purposive sampling: involves choosing members of a sample “for a particular purpose”

(Leedy and Ormrod, 2010:212).

3.4.1.2. Sampling design for the study

Parahoo (2006) advocates that in order to select a representative sample that can be used to generalise the findings to the target population, a probability sample should be used. Thus, cluster sampling was used. Cluster sampling is the most appropriate as the research investigates e-commerce adoption in two geographic areas i.e. Durban and Pietermaritzburg. In addition, any SMME that belongs to any of the business sectors identified in the National Small Business Amendment Act of 2003 was eligible to participate in the study. This enhanced the representation of SMMEs from all business sectors. Samples were taken from locations that cater for the entire spectrum of SMMEs. In other words, in both geographic areas, samples were taken from places that have Small, Medium and Micro businesses (i.e. Central Business District (CBD)).SMME owners/managers were selected as participants in the study as it is well documented in Chapter 2 that the perceptions of the owner/manager about e-commerce

53 influence to a large extent its adoption. SMME owners/managers were selected based on their willingness to participate in the research.

Coughlan, Cronin and Ryan (2007) posit that the size of the sample is an important factor to consider in research. Small samples present the risk of being excessively representative of small subgroups (rather than the whole population) within the target population. Gay (1996) provides guidelines for selecting a sample size:

i) For a population that has less than 100 individuals, the whole population should be sampled

ii) If the population size is approximately 500, 50% of the population should be sampled iii) If the population size is approximately 1500, 20% of the population should be sampled iv) If the population size is greater than 5000, a sample of 400 participants is required

The population of SMMEs that operate in CBD in both Durban and Pietermaritzburg was estimated to be greater than 5000. Thus, a sample of 400 SMMEs owners/managers was targeted, which was equally spread between the two clusters i.e. 200 SMME owners/managers were selected in each location.The size of the SMMEs business was determined by the number of full time employees (captured through the questionnaire).

Questionnaires were hand delivered to SMME owner/managers at their business premises.

Requests were made for the respondents to complete the questionnaire immediately. However, in most cases the respondents were not able to complete the questionnaire promptly. Thus, a maximum period of two weeks was given to the respondents to complete the questionnaires and arrangements were made to collect completed questionnaires. The data collection phase took 5 months (from November 2011 until the end of March 2012). Data collection was extended due to some challenges. In some cases, the respondents had not understood the questions, thus the collection of questionnaires had to be postponed. In other instances, questionnaires were spoiled due to negligence. In these cases, questionnaires had to be re- printed and re-administered which caused delays in collecting data. The longer period of data collection was also due to the fact that data had to be collected in two geographical locations.

Leedy and Ormrod (2005) state that in most cases, data collected by means of questionnaires i) reflect the reading and writing skills of the respondents, ii) and sometimes respondents answer what they think the researcher wants to hear. To address these shortfalls, i) efforts were made to

54 address and/or clarify any queries that respondents had before and after the respondents answered the questionnaire. ii) The researcher did not interfere with the respondent whilst he/she was filling in the questionnaire in order to minimise any influence over the responses.

A total number of 360 usable responses from SMMEs were gathered at the end of data collection, which represents a 90% response rate.

3.4.1.3. Inclusion and exclusion criteria

1. Inclusion criteria: Owners or managers of businesses i) operating in either the Durban CBD or Pietermaritzburg CBD, ii) whose business falls under the description of an SMME based on the number of full time employees in the company (according to the South African Small Business Act of 2003) iii) whose business falls under any of the business sectors listed in the South African Small Business Act of 2003, iv) willing to participate in the study.

2. Exclusion criteria: All other categories of employees and all other businesses that do not fall under the description of an SMME (based on the number of full-time employees) were excluded from the study. Businesses that do not belong to business sectors identified in the South African Small business Act of 2003 were also excluded.