• No results found

CHAPTER 3: RESEARCH DESIGN AND METHODOLOGY

3.5. The Research Methodology

3.5.1 Sampling

Use of samples to obtain precise information about a population is an efficient technique that enables a researcher to make inferences about the overall population (Yu and Cooper, 1983). Salkind (2006) suggested that in order to have generalisable results of the study, the sample should be a representative sample of the population.

It is important that the sample characteristics will be the same as those of the population. That is the reason why, sampling is centeral to survey design. Researchers believed that the choice between probability and non-probability samples should be based on considerations such as

145

the nature of the research, relative magnitude of non-sampling, and degree of error tolerance (Malhotra et al., 2007). Convenience sampling is a type of non-probability sampling in which people are sampled simply because they are ’convenient’ sources of data for researchers (Lavrakas, 2008). It is a statistical method of drawing representative data by selecting people based on the ease of access and availability. Taking into consideration what Creswell (2005) has suggested that a convenience sample includes participants who are available and willing to be a part of the study. One of the most important issues about any type of method is how representative of the population the results are. The current research is targeted towards Pakistani university teachers. Pakistan was selected due to its importance to the researcher as it is the home country of researcher, it is a developing country thus research based here would fill the gap in existing research with reference to lack of COO research based in developing countries, also, it is an emerging economy and research based in this country would be highly valuable to international business and marketing field. The Pakistani university teachers were selected as the research population of the current research as they are educated and there was a lack of evidence of any other COO research used this population. It was interesting to see how educated and affluent university teachers behave as consumers with reference to COO-image. Also, it was convenient for researcher to collect data due to personal references in universities.

When the variability in the population is low and the population is rather homogenous, the use of non-probability convenience sampling is preferable, as the use of judgment may allow greater control over the sampling process (Malhotra et al., 2007). The current research has adopted convenience sampling approach, targeted towards the homogenous population of university teachers of Pakistan. The current research used convenience sampling as it was

146

least expensive and time consuming of all sampling techniques. The sample was accessible, cooperative, and easier to measure. As research suggested that convenience sampling is used where projections to the populations are usually not needed, in such studies, interest centers on the proportion of the sample that gives various responses or expresses various attitudes (Malhotra et al., 2007). The sample of 1509 university teachers was chosen and a 100% response rate was achieved.

As mentioned earlier, convenience sampling is a method of drawing representative date by selecting people because of the ease of their availability and access. Research outlined various strengths of convenience sampling, such as, simplicity of sampling, ease of research, it can be helpful for hypothesis generation and pilot studies, it can facilitate data collection in short duration of time, and it is cost effective (Tolmie et al., 2011). At the same time this sampling technique has few weakness as well, such as, it is highly vulnerable to selection bias, sample might not represent the population as a whole, high level of sampling error, and generalisability is unclear.

Due to the security conditions (bombings and terrorism) in Pakistan (2008 to date), it was hard to administer the questionnaire to general consumers, as people generally avoid being contacted by strangers. It was also difficult to ensure the availability of a large number of respondents within a limited time period. In order to have a significant number of participants - almost 10% (1500 out of total 14616) of the University faculty members in Pakistan (HEC, 2011) have been recruited for the current study, using a convenience sampling approach. Thus the faculty of universities in Rawalpindi, Islamabad and Lahore (15 Universities and

147

approximately 100 participants each) such as: Mohammad Ali Jinnah University, Federal Urdu University of Arts, Science and Technology, Quaid-e-Azam Univerisity, Foundation University, Bahria University, Air University, National University of Modern Languages, Fatima Jinnah University, Riphah International University, University of Punjab, Lahore University of Management Sciences, Government College University, Kinnaird College University, University of Central Punjab, and Lahore College University, have participated in the current study. The sample was chosen mainly based on respondents’ convenient availability. The sample represented the characteristics of educated and middle income population of the Pakistani consumers.

The current research has increased validity of the convenience sampling by targeting a homogenous population (university teachers with certain level of education and income). The effective and efficient use of convenience sampling strategy was appropriate to the context and assured that generalisations are as convincing as possible to as many audiences as is feasible.

The population of current research was educated. Due to their high level of education, they were capable to answer questions with full understanding and that is a reason of achievement of high response rate. Also, the anonymity was ensured to improve the validity of their responses. In order to improve the validity, researcher should try to assure that the respondents participate in the study and keep dropout rates low (Trochim et al., 2015). The research further concludes that a high response rate is essential to infer from results to the wider population, also if the response rate is low then there is a greater likelihood of sample bias

148

(Wilson, 2014). Thus the validity of the sampling method in the current research, was also improved by achieving 100% response rate. Pilot study was also used to increase the validity of the sampling method.