Chapter 3 – Methodology
3.2 Research strategy
Sekaran and Bougie (2011) mentioned three types of research strategies which are known as the quantitative, qualitative and hybrid research which is mixed of both qualitative and quantitative data. Further, “most of the empirical studies of celebrity endorsement are quantitative in nature” (Choi, Lee & Kim 2005, p. 169). Supporting this Newman and Benz (1998) and Cohen (1980, cited in Sukamolson, n. d.) also argued that, statistics and quantitative methods are more weighted to the social, behavioral researches. Thus author has chosen a quantitative strategy where this research results are based on consumers buying behavior and their responses.
Table 3.1 - Quantitative Vs qualitative study
Quantitative Qualitative
A qualitative research refers distinguishing actions in the form of quality, features and performance of activities (Punch, 2005).
Sometimes complicated since it may not be able to gather data in every aspect because information regarding customer attitude and satisfaction cannot be measured by the numbers (Sukamolson, n. d.).
Researcher is responsible for data gathering, measuring and analyzing where it can be entitles for statistical errors (Thomas, 2003).
Easy to handle large number of data using this method and chosen population is not possible to conduct interviews and as (Sukamolson, n. d.) “Positivism adopts a clear quantitative
approach to investigating phenomena” that would easier to achieve clear
objective solutions and
recommendations (Crossan, 2001, p. 46).
“Quantitative research is an empirical research where the data are in the form of numbers” (Punch, 2005, p. 4).
Qualitative methods give more comprehensive findings which is not that complicated and easy to analyse (Moll et al, 1992).
Using qualitative methods would household the knowledge where allowing researcher for open ended questions and interviews (Moll et al, 1992).
It is not time and cost effective to gather information from a large number of populations through interviews (Morgan, 1997).
Adopts a clear approach to investigating information regarding customer attitude and satisfaction (Sukamolson, n. d.).
3.3 Research approach
Sekaran and Bougie (2011) conclude four types of research approaches which are named exploratory study, descriptive study, hypothesis testing and case study. Normally the two main methods are case study and surveys for a maketing researches which carries huge sample size (Gable, 1994). Sekaran and Bougie (2011, p. 30) described a case study as a “problem solving technique” and a critical method for translating the dynamics. In addition benefits and limitations of using this model will be described in Table 3.2.
Table 3.2 - Benefits and limitations of Descriptive study
Benefits Limitations
Purpose of surveys that classify people into numbers where reduces the uncertainty and it matches with this research objectives (Malhotra, 2008). Hence, through questionnaires and
observation the survey will be able identify exact desired outcomes out the relationship between common courses which are closely alike (Gable, 1994). Direct comparisons between the
factors will be more reliable and accurate in statistically with the one shot data collection, when handling a huge population (Frazee, 2005).
Although often surveys are “snapshot” while the time horizon will be cross- sectional where the data will be gathered in just one point of time and researcher may not be able to find every data according to his/her interest (Frazee, 2005; Gable, 1994, p. 113). Questionnaire should fully accurate
and cannot be amended after collecting the data and that is why a pilot survey is needed in surveys hence, researcher has pay to adequate time and money on these types of researches (Malhotra, 2008).
Researcher is responsible accurately in depicting the results (Thomas, 2003).
3.4 Research methods
As a research method close ended questionnaires will be distributed into predictable age group as to collect quantitative data for the survey. Questionnaire will be consisting with the simple dichotomy questions which discussed in Table 3.3. (Refer the Appendix 5 for the questionnaire). Although due to the large population the close ended questionnaire is easier to have a comprehensive and quick feedback. Time period for the research is 3 months.
Table 3.3 - Benefits and limitations of using close ended questionnaire
Benefits Limitations
Likert scale and multiple choice questions which discussed allow more reliability, accuracy, and validity (Park, Loomis, & Creel, 1991).
Five point Likert scale questionnaire which has the 5 options consisting strongly agree, agree, neutral, disagree and strongly disagree (Hartley & Betts, 2010), and multiple questions where people can choose one answer from the given multiple options (Kaplowitz, Lupi & Hoehn, 2004).
Due to the large population the close ended questionnaire is easier to have a comprehensive and quick feedback, (Akbayrak, 2000).
Time and cost effective since specifically scaled questionnaire limits the unnecessary data gathering (Baren
Questionnaires may assure uncertainty of the respondent’s truthfulness and they may be misunderstood about some areas (Akbayrak, 2000).
As an overall accuracy and reliability and validity of the questionnaire answer can be depended on their moods, behaviors, emotions, work and lot more factors (Lerner et al, 1997).
Can gather only limited data due to the specific scales and frames compare to the interviews and open ended questionnaire gathering (Baren et al,2003).
Less chance to researcher to gain new knowledge since questions and answers may set in predictable way (Ward et al, 1996).
et al, 2003)
(Author Developed, 2013)
3.5 Sampling approach
The sampling technique that will be used in this research will be probability sampling, as Sekaran and Bougie (2011) noted that this where usually the probability is known and equal for every person being selected within the population. The sample size is 100 and simple random sampling will be used out of “2,323,826 population” of Colombo district (Census of population and housing, 2012, n.p). Typically simple random sampling is supporting to quantitative data gathering as to reduce time and money wasting since this will be used to gather the data where processing entire data is not necessary (Wolman, 1954). Further errors and all other estimates can be measured in this method easily and the limitation is sample has to be assumed the population is free from different characteristics (Olken & Rotem, 1986).
3.6 Data collection
For the purpose of primary data this research will be carried out questionnaires comprised of close-ended questions including multiple choice, lickert scale and simple dichotomy questions. The secondary data will be gathered using the internet, books, journals, newspaper articles, and past researches, which has been published both locally and internationally
3.7 Pilot Test
The pilot test is a pre study to check the limitation of questionnaire and intended to plan the large scale of investigation with potential to assess the “feasibility study” (Thabane
et al, 2010, p. 1). With the prepared questionnaire, using the expected age group which is age 13-40, the pilot test was done within 10 respondents.
The first problem was found as there was no specific way to compare the two companies that considered in this research and to compare local and foreign celebrities that they use. Almost 5 respondents from the pilot survey have been asked “is this a regarding local celebrity or foreign celebrities?”. That appeared the importance of including my second question which intended to divide the whole sample into two groups which are interested in local celebrities with local cosmetic company and the people who are interested in foreign celebrities with the use of foreign cosmetic products. Finally the researcher found that question number 14 and 17 which were simple dichotomy questions where they had the choice of “Yes/No” and three people answered the question has mentioned “not always” and that assistance was helped out to put those questions in a likert scale method were researcher was able to collect more accurate data. (Please refer the appendix for the pilot survey questionnaire). Due to those changes that were made in the pilot survey tended the researcher to conduct another pilot survey to check the validity of the changes and that second one was successfully ended with making the final questionnaire for the survey.
3.8 Data Analysis
Walliman (2011) describe the importance of statistical analysis in descriptive study which requires checking the accuracy of the data gathered from the sample and easier to understand the findings. In supporting to this Miles & Huberman (1994) state that quantitative data analysis will help in order to explain and represent the research finding in a meaningful way to describe the distribution and relationship among variables.
Using SPSS which refers (Statistical Package for the Social Sciences) 17.0 software the data will be analyzed demonstrating through graphs and pie charts in chapter 4 (Priyatno, 2009).