This chapter introduces the chosen methodology for this thesis, following the “Research Onion”, presented in Saunders, Lewis and Thornhill (2012, p. 128). The chapter presents the research approach, the methodo-logical choice and the research strategy, as well as the data collection techniques and the data analysis. Addi-tionally, sections about research quality and research ethics are provided followed by a presentation of the re-search process.
As indicated by the problem statement and the purpose, not much has been written about purchasing and BSRs in the furniture industry, especially not from both the supplier and the buyer perspective. As the thesis took off from an empirical observation where the au-thors explored a phenomenon within the studied situation, it seemed to call for an induc-tive approach (Saunders et al., 2012) or an abductive approach (Alvesson & Sköldberg, 2008). The problem statement further suggests that much has been written about purchas-ing and BSR management in other contexts. In such a situation, when the researcher has access to a wealth of earlier theories developed for another context, Saunders et al. (2012) claims an abductive approach to be more eligible. In the process of further investigating the experienced phenomenon, in this case BSRs in the furniture industry, it was systemati-cally compared with existing theories, which created an iterative learning loop (Kovács & Spens, 2005). As such, the authors have moved between theory and data in a freely manner during the research process, which thus can result in subsequent modification of earlier theory (Dubois & Gadde, 2002; Saunders et al., 2012; Alvesson & Sköldberg, 2008). The research approach falls between an inductive and a deductive one but due to the iterative learning loop, the authors argue that this study has followed an abductive approach, see Figure 3.1.
The identification and choice of research approach early in the process will ease the choice of the overall research design including the methodological choice, the strategy and the da-ta collection techniques (Easterby-Smith, Thorpe & Jackson, 2008).
To be able to fulfill the purpose of this thesis, the authors have chosen to employ mixed methods research, using both qualitative and quantitative research methods. As the first RQ aims to explore and describe how the furniture industry manages BSRs, an observa-tion, interviews and a survey were conducted. This course of action is normally called se-quential mixed methods whose aim is to further investigate the results of one type of method by employing another method (Jacobsen, 2002; Creswell 2009; Boeije, 2010; Saun-ders et al., 2012), in this case an observation and interviews followed by a survey. The au-thors’ intention was to, through the observation and the interviews, get an insight of the BSR management that later could be explored in a wider population to investigate the rep-resentativeness of the findings (Creswell, 2009). However, the observation and the inter-views are also considered important as empirical findings in addition to the survey. A sim-ple explanation of employing a mixed methods approach is that it enables a richer under-standing of the investigated topic by utilizing two different research strands complementing each other (Teddlie & Tashakkori, 2009). As the entire thesis revolves around the devel-opment of BSR management, the baseline, i.e. the current BSR management, is best built by employing a survey strategy.
To be able to get an understanding of current BSR management and highlight potential opportunities for BSR management development within the furniture industry at a holistic level, a survey strategy has been deployed. A survey strategy is often viewed as highly rele-vant for exploratory research (Saunders et al., 2012), and is, as such, expedient for the pur-pose of this thesis. In addition, data collection from a whole population is enabled (Lee, Fielding & Blank, 2008), which further supported the authors’ choice when matching the research strategy to the purpose. The survey strategy was also found appropriate as the sought information could only be provided by the responsible persons at each company, which Vogt, Gardner and Haeffele (2012) argue is the main criteria for choosing a survey strategy. Furthermore, the cross-sectionality of this study is supported by the survey strate-gy as it gives a snapshot of the current BSR management for the chosen industry segment.
According to Saunders et al. (2012), the deployment of the survey strategy requires initial efforts to ensure an appropriate design of the instrument. Therefore, the observation, in-terviews and the literature study were also used as a pilot study to enunciate relevant ques-tions for the survey instrument. In the following sections the employed data collection techniques are described in detail.
In addition to the three primary data collection sources, a literature study has enabled sec-ondary data collection. For RQ 1, all four of the data collection techniques have been used, while a literature study, together with the findings for RQ 1, has been used for RQ 2, see Figure 3.2.
The execution of each data collection technique deployed in this study is described in the below sections.
During the initial seminar organized by TMF, both producers and suppliers of the furniture industry participated and discussed perceived issues concerning BSRs, see Table 3.1. The authors’ main objective during this seminar was to objectively observe the discussions without any interaction. Before the observation, the authors and TMF acknowledged the purpose of the observers to all participants to make them comfortable with the situation. The observation can, as such, be characterized as a type of observation called ”Observer Non-Participant” (Teddlie & Tashakkori, 2009, p. 219) where the participants are aware of the observers’ purpose but the observers do not take part in the activity under observation.
The observed participants were divided into three groups consisting of six representatives, both producers and suppliers, where one participant was designated as group leader. The groups were formed in advance by TMF to ensure that each group contained a mix of ex-trovert and introvert people to create best possible discussions without any communicative barriers. The groups got three topics touching upon strategic collaboration to discuss:
What does the current ”reality” look like?
What development possibilities do you see within strategic collaboration?
What actions and activities would you like to engage in to develop and improve the collaboration?
The authors were observing one group each and took notes about the discussions. After-wards the groups reconvened and presented the main points discussed, which led to cross-comparisons between the groups. Any observation can generate several interpretations (Saunders et al., 2012), and therefore it was valuable for the authors to compare the input and interpretations from three observed groups. The authors compiled the main points in a PowerPoint presentation that was kept as support for the thesis.
In order to get a general understanding about BSRs and to plot what has already been writ-ten within the subject, the authors first conducted a brief literature review. The literature review was thereafter extended into a literature study to create the theoretical framework. By taking this approach the current study can be connected to the constantly developing theories of the subject (Blumberg, Cooper & Schindler, 2011; Marshall & Rossman, 2006) and it is an attempt to fill the identified gap in the literature (Denyer & Tranfield, 2009; Marshall & Rossman, 2006). The literature study, which is concerned as a highly appropri-ate source of information (Olsson & Sörensen, 2011), was based on academic articles, books and dissertations and was conducted in an iterative process divided into five phases, see Figure 3.3.
In the first phase, the authors identified and determined four keywords related to the topic in consultation with Jenny Bäckstrand, Assistant Professor of Jönköping University, based on the observation:
Swedish furniture industry
An initial search was thereafter conducted in scientific databases accessed through the li-brary of Jönköping University. The search resulted in several hits, which were subsequently prioritized based on key authors and number of citations. As the articles were read and cat-egorized in suitable groups, new insights and refined keywords occurred whereby the au-thors returned to the second phase for additional searches. After a thorough search, prolif-eration of results became saturated.
To ensure the quality of the used articles, the authors have constantly tried to validate the information given in one article with another as well as choosing only well-cited articles, as it is imperative to have a critical approach to all information (Easterby-Smith et al., 2008; Creswell, 2009). By following a consistent approach in theory selection, measurement bias can be avoided (Saunders et al., 2012). The literature study has complemented the other da-ta collection techniques, but also provided interesting areas to investigate closer in the in-terviews.
The authors have chosen to use semi-structured interviews as one of the data collection techniques for the first RQ, as interviews are highly appropriate for accessing information (Blumberg et al., 2011; DiCicco-Blom & Crabtree, 2006; Easterby-Smith et al., 2008). As described in the previous sections, the interviews aimed to provide insights about the BSR management, where semi-structured interviews can be used to generate rich information about the subject of interest (Easterby-Smith et al., 2008), and is appropriate for explorato-ry studies (Blumberg et al., 2011). A semi-structured interview is characterized by a number of predetermined questions of open characteristics guided by the interviewer (Darmer, 1995; DiCicco-Bloom & Crabtree, 2006; Easterby-Smith et al., 2008) and allows probing of answers by follow-up questions (Darmer, 1995; Saunders et al., 2012). As such, the authors have been able to steer the interviews in directions to ensure access of relevant infor-mation. Concurrently, the open nature enabled the interviewees to freely give their perspec-tives of the subject.
In this study, the observation along with the literature study also provided a foundation for key questions and themes covered in the interviews. According to Saunders et al. (2012), the interviewer normally has some key themes that need to be encompassed, but these themes should be open for adaptation during the interview process. With this in mind the interview guide, see Appendix 1, was established based on the recommendations provided by Creswell (2009, p. 183) and Blumberg et al. (2011, p. 389).
As the companies in focus are located in Sweden, the interviews have been conducted in Swedish to facilitate the communication and to reduce the possibility for misunderstand-ings. Three of the four interviews were conducted face-to-face whereas one was made via telephone due to accessibility, see Table 3.2. Whereas face-to-face interviews are viewed as most suitable when the researcher is exploring a subject, telephone interviews can be used to improve the access to interviewees. The disadvantage of conducting telephone inter-views is often coupled to trust issues, where a low level of trust can mitigate honest an-swers to sensitive questions with an outcome of lower reliability (Saunders et al., 2012). In the case of the telephone interview conducted, the trust issue could to a large extent be avoided as the interviewers and the interviewee met and familiarized during the seminar provided by TMF.
The overall requirement for the interview sample was that there should be an equal distri-bution of producers and suppliers. The aim was, at first, to conduct four interviews with each type of actor. In consultation with TMF, producers and suppliers highly contributing to the discussion during the observation were, together with a few additional companies within the industry segment, selected as potential candidates. In addition, geographical proximity was a contributing factor to the selection. According to Boeije (2010) this kind of sample selection can be referred to as purposive sampling. However, during the process of sampling it proved challenging to access more than two suppliers and, therefore, to get an equal distribution, two interviews with each type of actor were conducted.
All interviews were recorded using the recording application of two smartphones, where one of the interviewers was guiding the interview while the other was taking notes. The re-cordings were later used to transcribe the interviews that were sent to each interviewee. This approach gave each interviewee the opportunity to rectify and clarify any misunder-standings as well as approve the interview. The process of correcting errors, i.e. data clean-ing, is of high importance for the research quality (Saunders et al., 2012). Based on the data collected in the pilot study, the results were also used to create the survey.
To get a general picture of the BSR management, a cross-sectional survey has been distrib-uted, see Appendix 2, to member companies of TMF representing both producers and suppliers within the industry segment, see Table 3.3. One of the major issues when using a survey as a data collection technique is derived from the sampling procedure, where the re-searchers must ensure that an appropriate sample is chosen (Creswell, 2009). In this study, the whole population, i.e. the member companies of TMF within the furniture industry in-cluding first tier suppliers, were used as a sample, which in total resulted in 72 companies. The member directory simplified the access to the respondents as it allowed the authors to create the sample directly, which can be referred to as single-stage sampling (Creswell, 2009). Before the sampling procedure, the authors decided to use the stratification in terms of company type, i.e. producer or supplier, to get both perspectives of the current BSR management. One of the main strengths of using a survey for data collection is that it ena-bles wide reachability of respondents (Easterby-Smith et al., 2008).
The process of creating and distributing the survey contained several phases, see Figure 3.4, where each phase hereinafter is described. For four of the phases, additional tools and in-put have been used, see the lower squares in Figure 3.4. Based on the collected data from the pilot study and the literature study, the authors generated potential survey questions, which were subsequently arranged into two main groups depending upon whether they were formulated from the standpoint of the supplier or the producer. The intention was that the survey should start with general classification questions, including e.g. type and size of company. This type of introducing questions enabled the opportunity to e.g. differenti-ate between producer and supplier respondents. After the introducing questions, the survey was divided into two parts where the questions presented were dependent upon if the re-spondent was a producer or a supplier. The questions were both of close-ended and open-ended characteristics where a 7-point Likert scale was used for the majority of the close-ended questions. However, a few ranking questions were also used where the respondent had to rank the importance of the available alternatives.
Table of Contents
1.2 Specification of Problem
2 Theoretical framework
2.1 A brief overview of the furniture industry in Sweden
2.2 Buyer-supplier relationships
2.3 Supplier selection and marketing
2.4 Portfolio approaches
2.5 Summary of theoretical framework
3.1 Research approach
3.2 Data collection
3.5 Research quality
3.6 Research ethics
3.7 Research process
4 Empirical findings
4.1 General view of furniture industry
4.2 Supplier perspective of furniture industry
4.3 Producer perspective of furniture industry
5.1 Current BSR management
5.2 Further development of BSRs
6 Discussion and conclusions
6.1 Theoretical implications
6.2 Managerial implications
6.3 Discussion and final reflections
6.5 Further research
List of references
GET THE COMPLETE PROJECT
How to develop the buyer-supplier relationship management An investigation of the Swedish furniture industr