This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
The previous chapter focused on drawing relevant academic and research on the dissertation research area. This chapter focuses on revealing the research design and methods adopted for the dissertation. It encompasses various areas including discussion on methodology and design of research, basis of selection of population and subsequent extraction of appropriate sample, strategy for data collection and description of the survey instrument adopted for this research and piloting of the same. It also reveals various aspects related to reliability and validity of scales adopted for this research. Finally, a brief overview of data analysis is presented.
3.1: Research Methodology:
Observing something, reading a newspaper to get information about new job offering or simply browsing the world wide web for new cell phone rates, all of these are form of research. Yet, it is important to mention here that it is used in quite different context in scientific perspective. The research problem posed in this dissertation asserts a scientific investigation so as to resolve the associated issue. The main rationale for adopting a scientific method to resolve the dissertation problem statement is to adopt a methodical approach towards the matter. This is consistent with the definition of research proposed by Webster (1985). It asserts that researchers should exercise due care and diligence and plan a systematic, step by step and coordinated endeavor so as to find and analyze the relevant facts and issues. The noteworthy outcomes of such an attempt is the furthering of existing paradigms and/or defiance or alteration to the existing body of knowledge. Furthermore, another way of approaching the concept of research is accumulation of dispersed information or statistics about a particular subject matter (could it be a living organism, non living organism or even any particular incident) brought under investigation.
Generally, research processes encompasses a series of same or similar steps which most of the times follow a rigid process protocol. However, depending on the nature of investigation, the researchers may alter the processes. Irrespective of the type of scientific inquiry, all researches require certain form of interpretation. Research could be conducted in actual field or it might be restricted to desktop or office of the researchers.
3.2: Design of Dissertation Research:
This section would describe the types of research conducted for the dissertation. On a broad spectrum, this dissertation involves both aspects of research which are qualitative and quantitative dimension. Adopting the methodology proposed by Zikmund (2000) and Cooper and Schindler (2006), the first was the crystallizing the management dilemma hierarchy. It involves series of improvement in the management question which is generally referred as problem statement. This is consistent with the advice of Zikmund (2000) who argued about the general inattention being given to this issue. Once the problem statement was crystallized, the next step was the conduct of a through literature review. For the purpose of the dissertation research, the areas of academic literature brought under review include customer satisfaction, service and service quality, online banking industry and banking industry of Pakistan. Based on this, the associated theoretical framework was established. This framework served as a guide in the whole research process by giving a path which must not be left in any case. Grossly, the review of literature brought various insights and relevant researches in the domain of dissertation problem statement. These insights served as the basis of inclusion of various constructs and associated variables in the research.
This dissertation adopts the research typology proposed by Cooper and Schindler (2006). They divided the scientific investigation into quantitative and qualitative research. The quantitative research is further categorized into descriptive as well as experimental research. Descriptive research is among the most popular form of research and is generally used to address questions about the reasons linked the occurrence of certain phenomena. It also addresses issues, for example, what is the type of information which is available with respect to some variable or other issue. On the other side of the continuum lies the experimental design. This is the form of the research design in which surveyor bring any desired type of adjustment so as to observe phenomena of interest under certain conditions. This change is then measured and is brought to use so as to generate meaningful insights about the subject matter. In contrast to experimental research design, various forms of descriptive research encompasses case studies of the organization or individuals, case cross-section along with retrospective research endeavors which are case cross-sections.
The typology of Cooper and Schindler (2006) shared various dynamics of research design. These include the level of crystallization of research issue at hand, data collection technique, the capacity or willingness to change any variable attached with the problem to be resolved, the overall purpose of research, time-scale, the research environment and how the respondents feel about the research endeavor influencing their day to day life. Based on the above mention framework, the current research endeavor can be regarded as exploratory as to the best of my knowledge and belief, no such research has already been conducted in the online banking context of Pakistan. The research tapped the responses on structured questionnaire which make it a communication research. Furthermore, as no change was posed in the environment of research, so the type of research on this continuum is ex-post facto design. Similarly, it is an statistical research as collected data would be summarized and various inferences would be drawn from it. The collection of data at one point of time (instead of collecting it at various points of time and that too for comparison) makes it a cross-sectional research.
For effective problem resolution, the research relies on both primary and secondary sources of data. Over the period of time, during different stages of research, various techniques were applied. These involve reference to secondary data including books, research articles, government documents, online magazines and in some cases the web pages. A proper reference (using Harvard referencing system) has been made in the reference section of the dissertation. In addition to these, primary data was also sorted for the dissertation. Zikmund (2000) revealed the advantages of collecting secondary data for being collected easily, faster and usually involving relatively less cost. However, the associated disadvantages are always there. These include currency and relevancy of data.
In the area of primary data collection, the methodology suggested by Wilson (2006). Both Wilson (2006) and McDaniel and Gates (2007) encouraged the use of survey for empirical research. This is because of the fact that all the respondents are exposed to same set of questions and surveyors have limited intervention power. Moreover, survey is also less costly, can be conducted in less time and tends to efficient in terms of resources.
3.3: Population and Sampling Frame:
This section deals with discussion of definition of population and sampling design. According to (Webb, 1998, Webster, 1985), for any particular research issue, the term population is used to refer to a collection of people, items or matters to which a researcher refers with an aim to extract a subset of the population. This subset, commonly referred to as sample is then put to evaluation or measurement. Extraction of a sample is influenced by a number of factors. Depending on the nature of inquiry, if the population encompasses individuals, then researchers extract a sample which comprises of human beings or people. These are generally referred as respondents and information is to be collected from these. Surveying sample instead of population has its roots in inferential statistics. Researchers use inferential statistics to estimate population characteristics based on their work on the extracted sample. In this case, the insights generated from studying the sample are generalized on entire population. Cooper and Schindler (2006) exerted that it is the studying population through studying sample which paves the basis of resolution of research problem. Furthermore, Grimm (1993) noted the qualitative characteristics of population which asserts that selected sample has to have a considerable representation of the research population. It means sample should have also characteristics which were thought to be present in the population. Furthermore, Grimm (1993) and Cooper and Schindler (2006) shared the benefits of targeting sample instead of focusing on entire population. Some of these include less cost, completion of research endeavor in time and saving from fatigue of accessing all the elements of the population.
3.4: Population Definition and Sampling Design:
For the current academic endeavor, all those persons who are currently engaged in using the online banking are regarded as population for the research. There are as such no restrictions in terms of age, gender, qualification, marital status, geographical area, etc. further to this, the target population involves people, who despite using online banking services, visited their respective banks' branches for any reason. These branches were located in Multan, Karachi and Peshawar. All three of these are large metropolitan cities of Pakistan. It is important to mention here that despite enormous penetration of online banking in Pakistan; still a significant portion of population does not have access to internet and/or are not well educated so as to enable them to handle the peculiarities of internet transaction.
Grimm (1993) and Mason et al. (1988) highlighted variety of sampling techniques yet put more emphasis on the two which are non-probability based and probability based. In the former, chances of selection of all the elements or cases of population are not equal whereas this is reverse in case of probability based sampling.
This research adopts non-probability based sampling method for its benefits. The major benefits include reduced cost, relatively few members of sample and elimination of cumbersome sample selection procedure. In non-probability based sampling, the researcher adopted convenience based sampling. The purpose of adaptation of this method was two-fold. Firstly, this method is cost efficient and secondly, researchers do not have to spend time in finding the potential respondents. Beside all these advantages, Tansey (2007) exerts that use of convenience based sampling method requires a word of caution.
3.5: Sample Size:
This section reveals discussion on the sample size of this dissertation. A sample of 300 bank customers was selected from 3 cities of Pakistan. The choice of cities was done judiciously. These cities are from 3 different provinces of Pakistan (out of total of 4 provinces). These cities represent large Metropolitan cities of Pakistan and enjoy vast population size and diversity and have a reasonable societal and cultural harmony. Furthermore, Karachi and Peshawar are also provincial capitals. The questionnaires were handed over to the members of customer services department of the branches. The customer services representatives approached the clients in the bank's waiting lounge. Since, respondents have to wait there and were not in a rush so quality of responses was deemed appropriate. Since, it was a non funded research i.e., the cost of conducting the research has to be borne by the dissertation cost, so some reference was made to the economy of research. However, it is imperative that such economy was not achieved at the cost of reliability. Generally, these two variables (reliability and economy) are on the opposite sides of the continuum. Both of these are desired characteristics of any research (Cooper and Schindler, 2006) so an appropriate harmony has to be yield between these two.
While conducting the survey, a special reference was made to sampling error. It is the probability or/and any particular orientation or prejudice which influence survey results. It also adds to the inaccuracies to data collection and the onward stages of the research. A relatively large sample size (300 respondents) was selected so as to cope with random error of the research. Access to funds and the amount of time available to complete dissertation were two important considerations in deciding the sample size. Yet, the number of good questionnaires collected from the branches was sufficient enough so as to meet the requirements of principle component analysis.
Factor Analysis is a multivariate data analysis which is used for data reduction. It is a statistical technique which is widely recognized for its use to reveal variance for a given set of variables with reference to relatively few factors (generally a group of variables). Such observed factors basically form a linear model with each other. The phenomenon has been widely adopted in behavioral sciences domain, including investment analysis, production operation management, marketing, sales, consumer behavior, etc. (Thurstone, 1931, Parasuraman et al., 2002, Litterman and Scheinkman, 1991, Paulhus, 1984, Costello and Osborne, 2005).
The work of Costello and Osborne (2005) and Garson (2009) revealed some of the recent insights in using the component factor analysis. Particularly, Costello and Osborne (2005) regarded it a multifarious technique couple with associated peculiarities of inferential statistics and data collection issues. They further underscored the widespread adaptation of varimax rotation and Kasier-Meyer-Oklin measure of sample adequacy. Yet, they also shared their concerns about the use of these two techniques. Similarly, the contribution of Costello and Osborne (2005) and Garson (2009) in the area of deciding about appropriate sample size to be used in conducting factor analysis. Garson (2009) highlighted various rules of thumbs. They pointed out towards a highly acclaimed amount of sample size for use of factor loading which is more than 50 subjects. Similarly, there are other rules of thumb which would be discussed in the following chapter.
3.6: Research Instruments:
The questionnaire used for this research has been adopted for earlier research. This research adopts the model as well as questionnaire proposed by Khan and Mahapatra (2009) who identified seven dimensions reliability, accessibility, user-friendliness, privacy/security, efficiency, responsiveness, fulfilment. The questionnaire has been widely used in previous research. The questionnaire had two sections. The first part of questionnaire dealt with the demographic and other general orientation about the survey whereas the second part of the questionnaire dealt with the actual analysis of perception of bank clients about the services offered by the banks. The questionnaire started with an introduction of the researcher and the respective university. It also narrated the objective of the dissertation which is a modified form of the dissertation's problem statement. The second part of the questionnaire was primarily based on the SERVQUAL model of Parasuraman et al. (1985). The variables used to measure the SERVQUAL model were grouped into various constructs. The constructs were exactly the same as proposed by Parasuraman et al. (1985). Most of the responses on questionnaire were tapped using a five point likert scale. Part two ended with an open ended question. The purpose of this question was a courtesy on the part of surveyor towards respondents so as to ask the later to share anything which was not mentioned the questionnaire.
3.7: Pre-testing of Research Instrument:
The questionnaire adopted for this research was pre-tested to check the content and language for the local targeted respondents. It was found that the language of the research was quite up to the mark. This was because of the fact that questionnaire has already been tested in many other researches across the globe and was even used in the same industry. The questionnaire adopted for this research was tested on 6 respondents and was then subject to pilot testing.
3.8: Field Work for the Survey and Response capturing:
In order to resolve the problem statement, the data was tapped from variety of sources. Some of those sources were from primary data sources yet secondary data sources also constitute a significant portion of this dissertation. For the purpose of gaining an understanding of already published material, a detailed literature review was conducted in the area of customer satisfaction, service quality, SERVQUAL, electronic banking and banking industry of Pakistan. For the purpose of this, various published scholarly peer reviewed journals, Government document, web page, online journals and magazines were subject to review.
For the purpose of primary data collection, the dissertation candidate contacted the customers' services representative (CSRs) of various banks across the targeted cities of Pakistan. The dissertation candidate shared the purpose of conducting research with the CSRs and convinced them about it. The blank questionnaires were then handed over to them and asked to get those filled from the respondents. As the customers were approached in banks and that even in those times when customers or clients do not have much to do, it not only gave convenience in collecting data but has also decreased respondents biasness as respondents have enough time to respond the questionnaires. As the customers were quite educated so questionnaire was not translated in native language (Urdu).
3.9: Validity and Reliability of Data:
The word of Miles and Huberman (Miles and Huberman, 1994) highlighted the importance of issues associated with reliability and internal validity of the research and regarded them of vital value. Validity of an instrument is the ability of the instrument to measure what it is meant for. As the questionnaire is adopted from previous researches and has been widely used so one can comment on the internal validity of the instrument. Similarly, reliability can be observed by computing Cronbach's Alpha of the questionnaire (Cronbach, 1951).
3.10: Data Analysis:
All the questions in the questionnaire were coded and were entered into Statistical Program for Social Sciences (SPSS), version 16. While entering data, each questionnaire was allotted the row number of SPSS. This is in order to identify which questionnaire goes into which row. For data processing MS Word was used and EndNnote® X1 was adopted for reference entry and editing. Factor Loading using varimax rotation and Kaiser technique to extract factors.
3.9: Limitations of Dissertation Research:
In addition to all the good things in the research, there are some issues which must be highlighted for the informed readers. These issues are some of the potential limitations of the dissertation. The findings of the dissertation would be based on the data collected from sample. Survey did not cover the whole population so the results reported in the dissertation should be used cautiously. Further to this, the use of convenience based sampling has its own disadvantages. Furthermore, the data was collected by the representatives of the banks. One should always discount it as there are associated issues of self reporting biasness. Moreover, the data collected from respondents may be subject to respondents biasness as they can submit incorrect answers. They might respond the questionnaire while they were in rush. The responses were available without any formal consideration so might not be correct. The findings of the research were based on only three cities of Pakistan. While generalizing the findings arising out of the collected, this factor must be kept in consideration. The questionnaire was adopted from some other research, despite all the goodness of the questionnaire, one should always be cautious while adopting some questionnaire. Most of the questions given in the questionnaire were close ended which might not reflect the original opinion of the respondents. This research did not select bank clients from the villages, hence missed lot of insights that may arise from them.
This chapter presented the research methodology of the dissertation research. It is an exploratory, cross-sectional, The questionnaire used was adopted from earlier research. This chapter also put forward issues related to population definition, sample selection strategy and sample size along data collection and analysis technique. Based on the data collected through field survey and applying the theoretical framework, the next chapter will reflect on the findings of the dissertation.
COOPER, D. & SCHINDLER, P. (2006) Business Research Methods, McGraw Hill.
COSTELLO, A. B. & OSBORNE, J. W. (2005) Best practices in exploratory factor analysis: four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10, 1-9.
CRONBACH, L. J. (1951) Coefficient alpha and the internal structure of tests. Psychometrika 16, 297-333.
GARSON, G. D. (2009) Factor Analysis.
GRIMM, L. G. (1993) Statistical Applications for the Behavioral Sciences., New York, John Wiley and Sons.
KHAN, M. S. & MAHAPATRA, S. S. (2009) Service quality evaluation in internet banking: an empirical study in India. International Journal of Indian Culture and Business Management, 2, 30-46.
LITTERMAN, R. B. & SCHEINKMAN, J. (1991) Common factors affecting bond returns. The Journal of Fixed Income, 1, 54-61.
MASON, R. D., LIND, D. A. & MARCHAL, W. G. (1988) Statistics: an introduction, Houghton Mifflin Harcourt P.
MCDANIEL, C. & GATES, R. (2007) Marketing Research Essentials, United States of America, John Wiley & Son, Inc.
MILES, M. B. & HUBERMAN, A. M. (1994) Qualitative data analysis: An expanded sourcebook, SAGE publications, Inc.
PARASURAMAN, A., ZEITHAML, V. A. & BERRY, L. L. (1985) A conceptual model of service quality and its implication. Journal of Marketing, 49, 41-50.
PARASURAMAN, A., ZEITHAML, V. A. & BERRY, L. L. (2002) SERVQUAL: a multiple-item scale for measuring consumer perceptions of service quality. Retailing: critical concepts, 64, 140.
PAULHUS, D. L. (1984) Two-component models of socially desirable responding. Journal of personality and social psychology, 46, 598-609.
TANSEY, O. (2007) Process tracing and elite interviewing: a case for non-probability sampling. PS: Political Science and Politics, 40, 765-772.
THURSTONE, L. L. (1931) Multiple factor analysis. Psychological Review, 38, 406-427.
WEBB, D. (1998) Segmenting police ‘customers' on the basis of their service quality expectations”,. The Service Industries Journal, 18, 72-101.
WEBSTER, M. (1985) Webster`s Ninth New Collegiate Dictionary. Meriam, Webster Inc.
WILSON, A. (2006) Marketing Research: an Integrated Approach, Essex, England, Pearson Education Limited.
ZIKMUND, W. G. (2000) Business Research Method London The Dryden Press.