Theory has a very important and indispensible role in the marketing research. Theory determines the variables, which are investigated and measured, and the source for interpreting findings. In order to develop theories, there are two approaches (1) deductive approach and (2) inductive approach (Bonoma, 1985). The deductive approach tries to find causality in the research area with the well-known theory in order to clarify the phenomenon (Malhotra and Birks, 2007). In other words, the deductive approach tests the theory by empirical evidence in the different settings (Perry, 1998). On the other hand, the inductive approach tries to find and identify an unknown research area, which does not have a theoretical background, by observations and in-depth questioning. Therefore, the inductive approach tries to build a theory (Bonoma, 1985).
According to Easterby-Smith et al. (1991, cited in Perry, 1998), the deductive and inductive approaches are founded from two different paradigms. The deductive approach is founded from positivist paradigm and the inductive approach is founded from phenomological paradigm. Moreover, the inductive approach consists of three different approaches: Critical Theory, Constructivism and Realism (Guba and Lincoln, 1994, cited in Perry, 1998).
Perry (1998) suggests that realism is more appropriate for case study research. To begin with, the inductive approaches should be used in the research areas, which are contemporary and in early paradigmatic stages such as relationship marketing (Boing, 1994, cited in Perry, 1998). Secondly, realism has more objective approach with respect to critical thinking theory and constructivism. Realism agrees with the induction approach that its purpose is to collect the data which is not observable in the first place. Moreover, realism accepts the interaction between the researcher and the phenomena and the personal involvement of the researcher (Hirschman, 1986; Perry, 1998). Thirdly, realism approach has its roots from the key theories. Therefore, realism does not have the disadvantage of positivism, which thoroughly depends on observable data, and critical theory and constructivism, which are incommensurable (Perry, 1998).
As it has been stated, realism is a type of inductive approaches with some deductive features. Realism targets to define and to describe the contemporary external world phenomena (Perry, 1998). However, it does not overlook the key literatures and theories. Therefore, realism cannot be stated as a pure deductive or inductive paradigm. Miles and Huberman (1994) state the importance of a prior research which may offer understanding about the subject in the research area. However, Miles and Huberman (1994) also state that this is valid in the research areas which need theory building. In this sense, realism seems to be appropriate for the case study research in this dissertation’s area of enquiry.
Rationale for Exploratory Single-Embedded Case Study Research:
There are several strategies to conduct social science research. According to Yin (1994), there are five main research strategies: (1)experiments, (2)surveys, (3)archival analysis, (4)histories, (5)case studies. Yin (1994) also states that these research strategies should not be assigned as exploratory, explanatory or descriptive purposes in nature. However, it is a fact that some strategies have certain advantages in different research purposes.
Case study research seems to be appropriate for this research enquiry because relation marketing, especially interpersonal relationships in Turkish manufacturing context, is existing and pre-paradigmatic real-life phenomenon which should be investigated in real-life context (Lindgreen, 2001). Yin (1994, pp.4) suggests that three conditions for determining the research strategy: (1) the type of the research question, (2) the extent of control an investigator has over actual behavioural events and (3) the degree of focus on contemporary as opposed to historical events.
There are five types of questions: (1)who, (2)what, (3)where, (4)how and (5)why. Yin (1994) states that “what” questions are mostly exploratory in nature. However, all of the five research strategies can be used in an exploratory study (Yin, 1994). In addition to this, “what”, “who” and “where” questions are appropriate for survey or archival record analysis (Yin, 1994). Furthermore, “how” and “why” questions are explanatory in nature and are applicable in the histories, case studies or experiments (Yin, 1994).
With respect to the control of investigator over actual events and the degree of focus on contemporary events, relation marketing, especially interpersonal relationships in Turkish manufacturing context is contemporary and cannot be manipulated because this research area is not appropriate to test in laboratory. According to Yin (1994), case study research is selected.
Source of Evidence:
There are six sources of evidence for the case study research: (1) interviews, (2) documents, (3)achieve records, (4)direct observation, (5)direct observation and (6)physical artefacts (Yin, 1994). In-depth interviews are direct personal interviews in order to discover new issues and insights (Malhotra and Birks, 2007). Interviews’ ability to uncover and explore the insight of complex phenomena is based on interviews’ seven features of knowledge: (1) knowledge as produced, (2) knowledge as relational, (3)knowledge as conversational, (4)knowledge as contextual, (5)knowledge as linguistic, (6)knowledge as narrative, (7)knowledge as pragmatic (Kvale and Brinkmann, 2009). The research interviews are the mediums where the knowledge is produced in the interaction of interviewer and interviewee. Therefore, knowledge is co-created by the questions of interviewer and the answer of interviewee. Production of the knowledge continues through the research stages until the conclusion. Secondly, the knowledge that is created through the interaction of the interviewer and the interviewee is relational. Thirdly, the knowledge is created through conversations which are about descriptions, narrations of cognitive notions, daily events and experiences. Fourthly, the knowledge is contextual because it created in an interpersonal context and sensitive to qualitative differences. Moreover, the knowledge is produced through language. Sixthly, Narration is a powerful tool because it consists of the experiences, the social realities and the interviewee’s own life. Finally, interviews have an unstructured or semi-structured approach. This allows interviewer to respond interviewees and execute useful actions which may provide more insight in the research area (Kvale and Brinkmann, 2009). Yin (1994) states that the interview is one of the most substantial sources of evidence for the case studies.
Yin (1994)’s case study approach is adopted for this research. Yin (1994) suggests a case study protocol which is a guide for interviewer and improves the quality of the case study. The case study protocol consists of case study questions. In this sense, the interviews are semi-structured. However, the case study protocol is a reminder of the important points for the researcher. Researcher may ask questions that are not in the case study protocol. Therefore, the interviews are semi-structured and open-ended. In the light of this, four case study protocols are constructed  (further detail will be provided in the research design section).
The Quality of Research Design:
It has been suggested that there are four tests that measures the quality of a social science method: (1) construct validity, (2) internal validity, (3) external validity and (4) reliability (Yin, 1994).
Construct validity is the test for creating accurate operational measures for the concept that are used in the case study. Moreover, the data collection should be objective. Yin (1994) offers three tactics to overcome the possible problems. These tactics are (1) using multiple sources of evidence, (2) the establishment of a chain of evidence and (3) the allowance of key informants to review the case study draft (Yin, 1994, pp.34).
Internal validity tests the fundamental relationships between constructs. In this sense, internal validity is the main concern of experimental and explanatory researches. In the light of this, this research, which is exploratory in nature, is not tested by internal validity.
External validity tests the accuracy of generalization of findings. This is a problematic issue for the case study research because case studies depend on analytical generalization. Surveys and experiments are based on statistical generalization. Replication of the case studies and surveys should give the same results. Therefore, multiple case studies are more appropriate for external validity (Yin, 1994).
Reliability tests that the consistency of the research procedures which should be followed by the researchers. Yin (1994) offers the case study protocol in order to ensure that the case study procedures are followed exactly. Secondly, developing case study database will increase reliability.
Single-Multiple-holistic-embedded case study design:
There are four possible case study designs. These are (1)Single holistic case study design, (2)Single embedded case study design, (3)Multiple holistic case study design, (4)Multiple embedded case study design.
Yin (1994) states three reasons to select the single case study designs. Firstly, single case study methods can be implemented in order to test and to challenge a well-established theory in the critical cases. Secondly, single case study can be used to explain a unique or extreme case. Thirdly, single case study can be used to research a phenomenon that cannot be observable or inaccessible. In addition to this, single case study method can be conducted as a pilot study for the exploratory enquiries.
Multiple case studies have certain advantages with respect to the single case studies. Herriott and Firestone (1983, cited in Yin, 1994) state that the multiple case studies are more robust. In order to conduct a multiple case study, the cases must be selected in caution that they should be similar however they also should be different with predictable reasons. In addition, performing a multiple case study needs time and resources.
The main difference between the holistic and embedded case studies is the focus of the case study. In holistic design, researcher investigates the case in a global-holistic approach without separating the individuals unit. In embedded design, the subunits of the cases study is identified and examined.
Due to time and resources constraints, single-embedded case study design is chosen. Therefore, the research investigates the business relationships and interpersonal relationships within these business relationships around a designated case study firm. However, each business relationships between the designated firm and its partners (suppliers and buyers) is identified as separate two subunits.
This research is exploratory in nature. Case study approach consists of eight in-depth interviews with respondents. In business context, it is a must to conduct interviews with both parts of the dyad in a relationship marketing research (Ambler and Styles, 2000; Zhang, Cavusgil and Roath, 2003). Furthermore, it is very substantial to conduct interviews with key informants in the case study firms (Harris, O’Malley and Patterson, 2003). Single embedded design for case study is appropriate for this research due to resource and time constraints. The case study consists of two subunits. First subunit investigates the relationships of the designated firm with its suppliers as a buyer. Second subunit investigates the relationships of the designated firm with its buyers as a supplier.
This case study research should be regarded as a pilot. Yin (1994)’s approach toward case study is more positivist in nature. Therefore, Yin (1994) suggests a semi-structured interview in which a research protocol is designed and followed by the researcher. Researcher has a list of questions in semi-structured interview that extra questions may be asked; however, the list must be completed. This approach provides reliability for the research. However, external validity is not sought for this case study because this research is a pilot case study.
The research is going to focus on a single firm and their relationship with its buyers and suppliers. Therefore; two interviews are going to be conducted with the key buyer and key seller from the designated firm. Afterwards, three interviews are going to be conducted with key sellers in three supplier firms which have relationships with the key buyer in designated firm. Same structure is going to be repeated with selling side of the designated firm with the key buyers in the buyer firms (See Figure X).
Figure X. Case Study Design
In order to do so, four case study protocols are designed for the key buyer in designated firm, the key seller in designated firm, the key suppliers of designated firm and the key buyers of designated firm. Therefore, this research targets the boundary spanners, who interact with each other during the normal flow of the business relationship. Thus, nature of key business and interpersonal relationships around the designated firm is going to be examined.
Initially, the designated firm ASP is contacted and a brief explanation of the research is provided. After agreement, a judgemental sampling is conducted in order to identify the suppliers and the buyers of designated firm. The key buyer and seller in the designated firm are identified and they consent to interview. The ASP is a SME. Therefore, the key decision makers are also the boundary spanners (This is also valid for the partner firms). The interviewers in partner firms are initially contacted by the ASP. Three interviewers from supplier side and three interviewers from buyer side agree to take part in this research. Eight face-to-face interviews took place in Istanbul and Ankara, Turkey. The interviews are between 40 to 100 minutes long (interviews are planned to be approximately 60 minutes). Interviews are conducted in the interviewers’ offices. A brief explanation and consent form is provided in the beginning of the interview. Interviewees signed the consent form and state their consent orally.
Data collection is a part of the data analysis. Data assembly is the next step. Firstly, the digital audiotape recordings are gathered and transcribed. According to Malhotra and Birks (2007), one of the challenges of the exploratory researches is the lack of fix questions. In other words, answers of the interviewee to a specific question could be different. Therefore, the data should be gathered into categories in order to access easily to the relevant data.
Coding is assigning labels a set of data (word, phrases, sentences or paragraphs) (Malhotra and Birks, 2007: Miles and Huberman, 1994). Moreover, memoing is a useful tool during coding because it enables researcher to take note about the data and to connect different pieces together (Miles and Huberman, 1994). As realism is the research paradigm for this research, codes are taken from the relevant theories and the key literature. The finding will be assigned according to the initial codes. However, new codes might be needed due to nature of the research. New codes will be created according to the findings. After the coding and memoing process, the data will be categorized and displayed with thematic categorization. Malhotra and Birks (2007) state that broad coding groups(themes) can be established after the coding. The themes are derived from the conceptual framework. However, as it is valid for coding, the new themes will be created according to the new coding.
Limitation of Case Study Research:
It is inappropriate to state that any type of research is free of limitations. Case study research has limitations. Yin (1994) identifies three points which are criticized by many researchers.
Firstly, case study research is criticized because there is a possibility of biased views to influence the direction of the research. However, this problem may occur in any type of research method. Yin (1994) suggests that researcher should treat any evidence fairly and should not concentrate on a particular aspect.
Secondly, case study research is ignored because the researchers concern that case study approach, especially single case study is not appropriate for generalisation. According to Yin (1994), this criticism about the single case study has a basis; however, same problem is evident for other research methods such as single experiment. This problem can be solved by conducting a multiple case study design. Moreover, he states that case study is not supposed to represent a sample. The aim of the case study research is to improve and analytically generalise theories.
The final concern is that case studies are time consuming. Moreover, the results of case study research are in big quantities that it is hard to conduct and analyse. Yin (1994) suggests that there are alternative ways of conducting and analysing the case studies.
In the end, all research methods have limitations. Yin (1994)’s approach to case study research tries to expel the main concerns and minimize biases. However, conducting a successful case study is hard to achieve.
Any form of research study creates vulnerability and so some ethical issues. Ethical issues are more important in in-depth interview because respondents are in more vulnerable position in interviews. Therefore, the research is going to be conducted according to Kvale and Brinkmann (2009)’s ethical protocol. Every stage of the research is critical in terms of ethical issues. Therefore, the Kvale and Brinkmann (2009)’s ethical protocol  will be the guide during conducting this research.
To begin with, this research should be beneficial to business-to-business relationship marketing. Therefore, the research gap is identified and questioned by the researcher that whether it is important to conduct this research.
Secondly, as the researcher is the responsible person of this research study, the researcher must not use the information, which is gained from interviews, for his own interest other than completion of this research.
Thirdly, researcher must protect the confidentiality of informants, their views and thoughts about specific subjects and persons. In this sense, this search needs special attention in interviewing and interpretation parts. Firstly, this research will investigate the business relationships between firms and interpersonal relationships between the participants. Therefore, both the interpersonal and the inter-organisational information are at stake. The researcher has responsibility to hide counterparts’ views, comments and thoughts from respective counterparts during semi-structured interviews. Secondly, the source of the views and the informants are masked throughout the research. The names of the participants and companies, that will be mentioned, are given by the researcher randomly.
Fourthly, the risk of involvement of this research is explained to the participants. Their oral and written consents are taken. Due to the agreement with the participants and their firms, all information will be masked and all data (interviews’ transcripts and digital recordings of the interviews) will be deleted after the completion of the research. Moreover, transcripts and interpretations of interviews are sent to the participants for their approval.