Introduction
The earlier chapter concentrated on international higher education, understanding the decision making process of Indian students and marketing strategies to further enable UoP to better understand the Indian market. This chapter provides a detailed overview of the methodology adopted for the purpose of examining the research objectives, method of analysis implemented for this research, a rationale of data collection method, techniques adopted to analyse the data, validity and reliability of the data and an in depth description of how each method was developed. Before investigating the data collected to justify the proposed arguments, the general research methodology shall be elaborated in this chapter to pave the way for a sound data interpretation and analysis. In the following text, first, research approach will be discussed in detail then followed by an evaluation of the effectiveness of the research methodology employed for this research.
Research Philosophy
Table
Multi-method approach to research
Method
Objective
Link to Research Aim
Sampling
Semi-Structured Interviews
To determine student’s views on major decision making factors while selecting and the channels used by them.
2, 3, 4
Purposive Sampling
Online Questionnaire
To determine a cross section view of students decision making process and compare it with semi-structured interviews and the literature review
2, 3, 4
Gatekeeper
Document Analysis
To compare it with the overall findings of the research.
1, 2, 3
NA
Case Study Approach
Introduction
Yin (p. 13, 2003) defines a case study as ” an empirical enquiry that investigates a contemporary phenomenon within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident”. The case study comprises of an all-inclusive method; it is a comprehensive research strategy. Keeping in this mind, Gummesson (p. 488, 2003) states ” that case study research provides an input of real world data from which concepts can be formed and propositions and theory can be tried”.
Yin (1993) has suggested that there are mainly three types of case study research
Exploratory,
Descriptive, and
Explanatory.
Researchers in business related subjects sometimes limit case studies to the exploratory use. For Instance, exploratory case study can be used as a basis for formulating questions or hypothesis testing. Whereas, descriptive case study attempts to describe, what happened to a product/service when it is launched. Lastly, explanatory research can be useful for example to study processes in companies. For this research, it seems exploratory case study fits perfectly, as this case explores past research on the subject, tests the hypothesis.
Why case study approach?
According to Yin (2003) a case study method is appropriate when: (a) the focus of the study is to answer ” how” and ” why” questions; (b) one cannot manipulate the behaviour of those involved in the study; (c) one who wants to cover contextual conditions because they are relevant to the phenomenon under study; or (d) the boundaries are not clear between the phenomenon and context.
For an illustration, there has been a similar study conducted on the decision making of nursing students conducted by Baxter & Rideout (2006) sought to determine the types of decisions made by nursing students and the factors that influenced the decision making, which is similar to this research where the researcher is probing Indian student’s decision making. A case study method is perfect for this research, because the case discusses the decision making of Indian students, but the case could not be considered without the context, the University of Portsmouth. It would have been impossible for the researcher to have a true picture of Indian student decision making without considering the context within which it occurred.
In order to obtain a strong understanding of the study’s requirements and since the research involves investigating UoP, which indicates this study as a case study analysis in the form of UoP. Therefore, Case study research will be an ideal method as understanding student’s decision making process. According to (Soy, 1997) adopting this method will only add depth to the research giving an extended strength to the past research conducted in this area. A research by Soy(1997) cites many well-known case study researchers like (Stake, 1995), (Simons, 1980) and (Yin, 1984) who have written about case study research and suggested techniques for organizing and conducting the research successfully. All of them seek to ensure that the topic of interest is well explored, and that the essence of the phenomenon is revealed, but the methods that they each employ are quite different.
This introduction to case study research draws upon their work and they have proposed six steps that should be used during a case based research. This research follows a similar pattern.
Determine and define the research questions
Select the cases and determine data gathering and analysis techniques
Prepare to collect the data
Collect data in the field
Evaluate and analyze the data
Prepare the report
Strengths & Weakness of Case Study Method
The hallmark of case study research is the use of multiple data sources; a strategy which also enhances data credibility says Patton (1990). (Baxter & Jack, 2008 ) says there could be many Potential data sources like: documentation, archival records, interviews, physical artefacts, direct observations, and participant-observation. Case study research is unique in comparison with other qualitative approaches, where a researcher can collect and integrate quantitative survey data, which further facilitates reaching a holistic understanding of the research being studied. In case study, data from these multiple sources are then united in the process rather than handled individually. Each data source is one piece of the ” puzzle,” with each piece contributing to the researcher’s understanding of the whole research. This convergence adds strength to the findings as the various pieces of data are then interlocked together to promote a greater understanding of the case.
Both Stake (1995) and Yin (2003) base their approach to case study on a constructivist paradigm. They claim that truth is relative and that it is dependent on one’s perspective. This paradigm ” recognizes the importance of the subjective human creation of meaning, but doesn’t reject outright some notion of objectivity” (Miller & Crabtree, 1999, p. 10). The advantage of this approach is the close collaboration and interaction between the researcher and the participant, while enabling participants to tell their stories (Miller & Crabtree, 1999). Through the medium of these stories the participants are able to describe their views in reality and this enables the researcher to better understand the participants’ actions (Lather, 1992).
While, Zainal (p. 2, 2007) has criticised ” case study method in terms of its lack of robustness as a research tool, thereby the need of crafting the design of case studies becomes a paramount importance. Researchers can adopt either a single-case or multiple-case design depending on the issue in question”. In our case the researcher is concerned with just a single case. Zainal (p. 5, 2007) has mentioned another drawback of a single-case; it is the inability to provide a generalising conclusion.
Triangulation
O’Donoghue and Punch (p. 78, 2003) define triangulation as a ” method of crosschecking data from multiple sources to search for regularities in the research data”. One way of overcoming the generalising problem, is by triangulating the study with other methods in order to confirm the validity of the process. Due to the contextual nature of the research, this project adopts a multi method case study approach, combining qualitative and quantitative approaches. As (Saunders, Lewis, & Thornhill, 2007) points out, multi-method allows different approach for different research purposes, since each method has its own strengths and weaknesses (Smith, 1975). Secondly, multi methods allow triangulation, thereby increasing the validity of research.
Multi Methods
McNeill (1990) states that the choice of research methods often depends on the nature of the topic, and the amount of time, money and work hours available. Many authors like Bryman,(1988), Easterby-Smith et al,( 2002) in the past have distinguished qualitative and quantitative research (Saunders et al, 2007) and also agree that arriving at a distinction between the two is often a dilemma.
Qualitative methodologies have had an extended tradition with organizational and management research providing strong tools for management researchers (Cassell and Symon, 2006). Quantitative methods may be used alongside these to provide additional scientific rigour. Tashakkori and Teddlie (p. 146, 2003) assert that by using multiple methods for analysis provide useful and facilitate a path to get an appropriate solution to the research question and evaluate the extent to which research findings can be trusted and inferences made from them. Whereas Gummesson (2003) states that neither using numbers (quantitative) or words (qualitative) in research is unimportant. Both requires correct interpretation and accurate research results depend on how the data is generated, analysed and finally interpreted
VALIDATION OF PAST RESEARCH
The review of existing research presented in Chapter 2 has encountered a critical problem. This was the lack of adequate information about decision-making factors influencing Indian students’ behaviour and use of channels while selecting UoP, as most of the related research in this field has been conducted by researchers from western countries. The first phase of research is aimed to achieve the above purposes. First step is to enable decision-making factors to be identified on the basis of their likelihood in relation to the research questions. Second step, to provide a base of further comparison of factors identified through the in-depth interviews and online questionnaire.
Data Collection Method
Assumptions of the Research
There are many other factors like family influences, cultural and environmental influences, that may influence the students’ intention to study at a university. However, those factors are considered insignificant, as the impact level may be negligible to the result of this research.
All the respondents were assumed to understand the questions in the survey, and answer them honestly. As a result, the findings in this study represent real situations.
The case in this context discussed is University of Portsmouth’s (UoP) Indian student community. A key strength of this case study method involves using multiple techniques in the data gathering process. Data gathering methods used:-
Figure 4. 1
Secondary Data
The initial phase of data collection involved the gathering of some previous empirical studies that data could be used with current research. McGivern (2006) explains that external secondary data from journals, articles and government sources could be used for current research purposes. Hence the research was mainly undertaken through desk based research and refers to academic articles published by various researchers in the field of marketing higher education and finally a research conducted by British Council (Refer Appendix I) about Indian student decision making process is employed for a comparison with the researcher’s results which will assist the researcher to come to a more reliable conclusion. The above use of secondary research enabled the researcher to critically evaluate past research whilst identifying trends and patterns that could support or dismiss theories developed from the primary research.
Semi-Structured Interviews with Indian Students
According to Bagozzi (1994), Interviews are one of the most widely used qualitative data collection methods within the marketing discipline. Primary research was undertaken through semi-structured interviews engaging prospective Indian students. It is defined by Malhotra, (p. 147, 2004) as ” an unstructured, direct, personal interview in which a single respondent is probed by the interviewer to uncover underlying motivations, beliefs, attitudes, and feelings on a topic”. The semi-structured interview lies between two extremes: – structured interviews which are characterised by pre-formulated questions and unstructured interviews which are characterised by more flexible questions (Kardorff & Steinke, 2004). The researcher chose semi-structured rather than structured interview because it offers sufficient flexibility to approach different participants differently. This was an exploratory study using a qualitative approach to understand students’ attitudes and opinions. The interview structure is created keeping in mind the academic literature and the subsequent gaps that the researcher would like to further investigate which is to discuss their motivations and reasons that made them choose UoP.
The purpose of doing semi structured interviews was to allow detailed understanding through discussions and gain insights into the decision making process of the students. In addition, according to Kvale (1983, p. 176), a good qualitative interview should have a lower degree of structure imposed by the interviewer, a predominance of open questions is an effective method, hence the researcher emphasized on asking some fixed questions to obtain relevant information and also had provision for the interviewees to elaborate on their own answers. Even Robson agrees that interviews are more effective than a questionnaire when open-ended questions are involved (Saunders et all, p. 281, 2003). Also, decision-making factors in selecting UoP can be both subjective, and objective, in-depth interviews will allow the researcher to search for the ‘subjective’ meaning of a specific issue, and then relate these meanings to an ‘objective’ social structure (Silverman, 2000).
Since the researcher has interacted with the participants, the study takes up an interpretivist paradigm in which knowledge is believed to be created by the researcher in a socio-cultural context (Cohen, Manion, & Morrison, 2000). This is also since the research objective of this study is to probe consumers (students) behaviour which requires deeper understanding of decision making process, interacting with the participants in an informal and personal language. It is clear that the goal of this research is to derive meaning, with the aim to recover and understand meanings/systematic divergences in meaning. Hence, this research uses the questioning, interpretive approach. According to Greenfield (2002) the benefits to an interview method is that it increases the comprehension of the data and makes the data collection more organised for each participant. However, the negative aspect is that there is a fear that important and significant issues may be accidently left out. To overcome this, interviewees were given as much time as possible to answer questions and any deviations from the main questions were explored thoroughly.
There were 20 interviews in total which were conducted with prospective students and each of them lasted for around 30 minutes. The interviewer’s role is critical to the success of this research. Malhotra (2004, p. 150) suggests that the interviewer should: first, avoid appearing superior and put the respondent at ease; second, be detached and objective, yet personable; third, ask questions in an informative manner; fourth, not accept brief ” yes” or ” no” answers; and lastly, probe the respondent.
One in- depth interview with a study abroad agent was also carried out to get their perspective of why did Indian students chose UoP. Interviewees were sent an e-mail introducing the interview process (Refer Appendix C), mentioning their rights to confidentiality, to withdraw from the interview at any time; to withhold information; and to informed consent. It also outlines data collection and storage methods.
Pilot Study of the Interviews
After an initial pilot interview with the first interviewee, (Refer appendix ) the researcher identified some problems which needed to be addressed and modified in order to produce the final structure.
To start with, the pilot interview revealed that some of the questions should be modified in order to prevent a simple yes/no response or even just a one line answer. Although the researcher tried to maintain an open approach with regards to the questions asked (Refer appendix) it did not always ensure that the answers would be similar. For example, Did you get easy access to the information you wanted?, the interviewee simply answered ” yes”. It then became very important to stress ‘how’ and ‘why’ in order to gain a more insightful answer from the interviewees. Over time, the researcher was able to develop a technique by asking follow up questions which are disguised as non-leading, but are in fact designed to delve further into a more relevant and thorough answer (Rapley, 2004).
The researcher during the pilot interview asked questions like ” why did you decide to study abroad”? Which was irrelevant for this research and therefore it was eliminated. Also to get a comprehensive feedback the researcher decided to ask the interviewees their valued recommendations.
Sampling Strategy
Purposive sampling was adopted by the researcher to select 20 prospective students, ranging from both the gender, for all the disciplines and at all level of studies. The sample was contacted by using the database of the agent, ILW, India based in Mumbai. The targeted sample was defined as such, as these groups of students have the highest possibility of continuing their study at a HEI abroad. In other words, their intention to continue their study at UoP was assumed. Almost all of the sample population was concentrated around major urban areas in Mumbai.
The potential interviewees were contacted and informed about the nature of the research project. Further, interviewees were asked for references of their friends/family that could be interviewed and who were appropriate for the research. This allowed a ‘cascading’ approach, whereby further willing participants were found via friends of friends. According to Rapley (p. 17, 2004) ” there are many problems of recruitment and when assessing potential participants, you have to follow many trails, often relying initially on friends and colleagues and then contacts given by other interviewees'”. Whilst purposive sampling comes under non-probability sample, Saunders et al (2007) states that it is more frequently used during a case study research and is perfect when working with informative or special subjects, like in this case.
There is also a major criticism often directed at studies employing this sampling approach is about how representative it can be (Rodham, 1998). However, Sayer (1984), cited in Rodham (p. 82, 1998) argues, ” Providing there is no pretence that the whole population is represented, there is no reason why an intensive study should be less objective about its particular subject than an extensive study”.
The sample list of willing participants can be found in appendix. The sample indicates that the majority of the students were interested in pursuing post-graduate studies, in the field of Business/Management fields.
Approach to Recording the Data
The dilemma on whether to record the interviews or not, was arrived at by considering Glaser, (1998) who recommends against using audio-recording, and Saunders et al (2007) who acknowledges the benefits and draw-backs of audio-recording interviews. (Saunders et al, p. 334, 2007) has pointed out some benefits: – ” accuracy, lack of bias and providing a permanent record, the disadvantage is the recording apparatus may adversely affect the relationship with the interviewer, the style of the interview and inhibit responses”. The researcher decided to record the data by taking notes as the interview progressed.
Once the interview was done, the researcher made sure, the details of every interview was noted on the same day, complete with assigning a code number for their future identification, so as to avoid any mix up of data with the other interview records, just as recommended by (Ghauri & Gronhaug, 2005). To maintain anonymity of the interviewees, all the interviewees were assigned code number for their identification.
Interpreting the Data
As the interviews were to be administered in India, with prospective students and, for an open conversation and to get as much information from, ” the researcher interviews in primary language which is not English; researcher and participants are ‘fully and fluently bilingual – they slip between two languages during the interview” (Rossman and Rallis, p. 161, 1998). In this case the other language used was Hindi, thereby raising the risk of losing crucial data while interpreting and translating it to English language.
Collecting data some of the data in Hindi and presenting the findings in English can have a direct impact on the validity of the research and its report. Since, the researcher acted as a translator for this research, Birbili( 2000) has cited Vulliamy (p. 166 1990) who has stated ” where the researcher and the translator are the same person the quality of translation is influenced by factors which can affect the quality of translation during the research: the linguistic competence of the translator’s; the translator’s knowledge of the culture of the people under study; the researcher’s fluency in the language of the write-up and finally the circumstances in which the translation takes place”.
There are some problems faced by the researcher while collecting data. Like for example, some words have existence in Hindi language but not in English, differences between the two languages like grammatical and syntactical structures are some of the major issues which demands specific decisions. These decisions along with the above factors have a direct impact on the quality of the findings of the research and the resulting reports.
(Birbili, 2000) has mentioned some useful techniques to overcome translation-related problems like back translation, consultation with people during the translation process and piloting the interviews. For this study, the researcher adopted the use of pilot study in order to negate translation problems.
Also while interviewing the participant, the researcher for a better clarification, particularly made sure to ask respondents not only for their answer but also for their interpretation of the item’s meaning to avoid wrong interpretation.
Structured Online Questionnaire
Structured questionnaires entail a series of closed-ended questions from which respondents choose their answers. This tool was employed because the researcher wanted to get a quantitative measure of channel choices when searching for information for UoP and the reason to choose UoP. This formed the first part of the data collection from the students to obtain background information that could be explored in the second part i. e. the semi-structured interview.
Sampling Strategy
The survey was conducted through a mode of online questionnaire facilitated through a gatekeeper in the form of the ” International Office”. As Saunders et al (p. 171, 2007) in his book mentioned that it is very important that before getting access to the information the researcher had to clearly mention the objectives of the research to gain confidence of the International office, which thereby lead to an access to the database.
Considering the time constraints of conducting interviews within a Master’s level dissertation, the researcher feels it is necessary to obtaining data via online questionnaire, in order to obtain a greater number and range of responses across the population. The responses will not be as thorough as the in-depth interviews but will at least serve to validate the findings.
The population under study was current students who had experienced any of the marketing channels or activities used by UoP. There was particular focus on Indian nationals. The sample did not have a balanced gender distribution and equal distribution of past and current students. In assistance from the International office, emails were sent out to prospective students requesting their consent to participate. The questionnaire was distributed to all the 111 Indian students on the campus which constitutes 100% of the population.
Questionnaire Design
The questionnaire was created using Survey Monkey online software package, the benefits being that it was self-administered over the internet, data input and responses were automatic with the participant’s name as anonymous. The strategy was to send out the questionnaires twice with a week’s gap to cover the maximum population.
Foddy’s (1994) understanding that when designing a questionnaire, the question must be understood by the participant in the way intended by the researcher and the answer given by the participant must be understood by the researcher in the way intended by the participant. As suggested by (de Vaus, 1996) & (Gillham, 2000) short sentences were used, negatives were avoided and questions were designed to be as precise as possible.
Basically, the questionnaire consisted of three sections:
Section 1: Respondents demographic information
Section 2: Rating important factors influencing decision to chose UoP
Section 3: Respondents choice of channel
The questions prepared were keeping in mind the audience and constructed in a way to prevent any ambiguity. Section 2 of the questionnaire measures the factors most important while selecting a university; it had 15 factors listed, whereas Section 3 questioned the channels used to search information.
Questions from Section 1 were related to demographics on gender, the department, and the level of the study, meant that the data could be analysed for differences between these. This helped in examining the student decision making process.
Responses in Section 2&3 were measured on a five point Likert-style scale, based on how far respondents agreed with each statement. A Likert scale is a uni-dimensional scaling method and chosen for this thesis. As a first step, the idea is to measure the importance of different factors when selecting university. The factors were scaled on a 1-5 response possible scale. The chosen numerals for the questions in our conducted questionnaires were from one to five and the chosen scaling model was a Likert scale. The numbers were presented in a ” Likert scale” format for ease of interpretation and the possibility of giving a ranking order. The number one is defined as of no importance to our respondents and the number five as decisively important. The numbers, as explained are mapped to the questions, i. e. the important factors giving the possibility so see a pattern as well as ranking (Ghauri & Gronhaug, p. 76, 2005). When measuring an object, scaling as a method in research gives a clear, visible result.
Pilot Study of the Questionnaire
Saunders et al (2006) in his book has cited (Fink, 2003b), suggested that to avoid participants having problems in understanding while responding to the questionnaire and also to monitor that there is no problem while recording the data, pilot study should be undertaken. In addition it will enable to assess the validity of the questions asked and also test the reliability of the data collected. The purposes of the survey and pre-testing were explained to the group of 10 students as similar to the final population from the sample. The sample size of 10 students is selected on the basis of the recommendation by (Fink, 2003b) from Saunders et al (2006).
Table 4. 2 Sample of Indian students participating in pre-testing of the survey.
Male
Female
Total
Post Graduate
5
2
7
Under Graduate
3
0
3
Total
10
All participants were asked to complete the draft questionnaire before they engaged in discussion about each question. This discussion was recorded and conducted in English in order to reduce the complexity of communicating and to enable participants to express their feedback more accurately. The purpose of the pilot test is to refine the questionnaire so that participants have no problem in answering the questions . The responses were gathered personally from each student and a feedback was collected with checks made on the responses and timings. The researcher double checked the pilot questionnaire to ensure that the participants didn’t face any problem understanding while answering the question and have followed the instructions correctly.
The feedback enabled the questionnaire to be improved through rephrasing certain questions and reformatting the questionnaire were respondents felt it didn’t have a smooth flow. The pilot test helped in corrections and modifications, for example while selecting an option for the question in the second section, there seemed to be a technical error with the software, as the participant could select multiple options (factors) in their decision making. This pattern was similar for Section 3. 1 and 3. 2. All were revised in the same structural manner.
Another potential mistake was identified in section 3. 2, here the participant had to tick the level of agreement, there had to be five likert scales of options to choose from Strongly Disagree, Disagree, Moderately Agree, Agree and Strongly Agree. But it seems one option Strongly Disagree didn’t appear in the questionnaire. Accordingly, the error was rectified. At the end, all the 10 participants were thanked for their inputs.
Overall the above process aided the researcher to find out how long the questionnaire took to complete, whether any questions were unclear or ambiguous, whether the layout was clear and attractive. The pilot test increased the reliability and validity, and made sure that the data collected was accurate and consistent.
Data Analysis
Interviews
The researcher anticipated participants to express their opinions and therefore open-ended questions were employed. Firstly, data gathered from the interview was transcribed to reduce the data complexity. The next staged involved careful and intensive reading of the data guided by some steps as suggested by Miles and Huberman (1994). This includes the counting of frequency of ideas from the responses while noting patterns and themes to facilitate the categorisation of the responses. Categorisation involved identifying key topics from the interview summaries. The aim was to note the individual aspects of recurring topics and relate it in a broad sense, to the context of the study. These included:
Information search
Channel Choice
Communication within channels
Type of information provided
Relevance of information provided
Decision Making Process
Questionnaires
Analytic Statistics such as mean, variance, standard deviation were calculated using SPSS software and also satisfaction ratings were calculated to measure the satisfaction of the information received through various sources and channels. Finally all the data was then exported into Microsoft Excel for further calcu
