Buyer and Supplier perspective in interfirm relations: The role of governance in the buyer - supplier relationship and its influence on ACAP and ultimately on learning performance

(1)

Buyer and Supplier perspective in interfirm relations:

The role of governance in the buyer - supplier relationship

and its influence on ACAP and ultimately on learning

performance

(2)

(3)

Master Thesis 2015| Margarita Racheva |3

Buyer and Supplier perspective in interfirm relations:

The role of governance in the buyer - supplier relationship

and its influence on ACAP and ultimately on learning

performance

Master Thesis, MSc Marketing Management,

University of Groningen, Faculty of Economics and Business,

Department of Marketing

22 June 2015

Margarita Georgieva Racheva

Student number: S2699923

Address: Oude Riet 40, 9746 PL Groningen

Tel: + 31 (0)616759059

E-mail: margarita.racheva21@gmail.com

(4)

Abstract

Different governance modes can influence absorptive capacity and ultimately learning performance. Therefore, many authors studied the links between these constructs in some of the most common business relations, namely those between buyer and supplier. Large number of the studies on this topic focused only on the buyer or supplier perspective, using information collected only from one side of the relationship. Therefore, to avoid inaccuracy in the results researchers focus on obtaining two-sided information. As a consequence, in the studies conducted using dyadic information the bias problem is resolved. However, the actual differences between buyer and supplier are rarely main focus of the studies and additionally, very often center of the research is the long term relationship with high value. This, in many cases, leaves not fully investigated the individual point of view of the buyer and supplier and additionally the short term relationships. These neglected conditions give space for further research, therefore the direction of this study is to fill to some extent these gaps. The present research investigates the effects of relational norms and contracting as governance modes on the two dimensions of absorptive capacity - potential and realised and ultimately the effect on explorative and exploitative learning performance. The statistical analysis is performed in SmartPLS3.2.1, because of the possibility to use smaller samples of data and additionally, because of the possibility to evaluate complex models including more than one dependent variable. In the program a Multi-group analysis and mediation analysis were performed to determine the direct and indirect effects between the constructs. Throughout this research we aim to study the different perspectives and associations of buyer and supplier and furthermore the effect that the duration of the relationship has on the viewpoints of the partners. The final results suggest that even when working together and having access to the same information, two firms may differ in their understandings, regarding certain aspects of the relationship. Moreover, the length of the relationship has been found to additionally influence the differences in the interfirm relations. At the end of the study are presented ideas for implications and possibilities for future research.

(5)

Preface

Making the decision to follow the path of Marketing was not an easy one given the countless opportunities that young people have today. Before making my choice I was hesitant whether or not this is the right career path for me. However, without doubt, now I can say that this decision was the right one. Moreover, I believe that the MSc of Marketing in the University of Groningen has helped me gain important knowledge on the topic of marketing and has provided me with valuable experience. The final part of my studies is the present master thesis on the topic of absorptive capacity.

Given the opportunity, I would like to thank the people who helped me and made the writing of this thesis possible. To begin with, I want to express my gratitude to my supervisor Dr. Hans Berger who has guided me through the confusion and struggles of the process and who has always been helpful and positive. Next, I want to thank everyone in my thesis group for the valuable insights and for the friendly environment that we created together. Also special thanks goes to all my friends, who helped me with suggestions and ideas for the realisation of this thesis. Last but not least, I want to thank my family who constantly supported me during the entire year and backed me up in my difficult moments.

Groningen, June, 2015

(6)

List of Tables

Table 1: Construct reliability and validity of reflective scales. ... 27

Table 2: Convergent validity of formative scales... 28

Table 3a: Squared roots of the average variance extracted (AVE) and correlation matrix. First order constructs. ... 29

Table 3b: Squared roots of the average variance extracted (AVE) and correlation matrix. Second order constructs. ... 30

Based on the results from the analysis, we can conclude that the model as an overall is strong and can permit the further investigation of the data. ... 30

Table 4: Standardized path coefficients ... 32

Table 5a: Multi group analysis results, buyer - supplier data. ... 33

Table 5b: Multi group analysis results, short - long duration ... 34

Table 6: Mediation effects ... 38

Table 7: Hypotheses ... 39

List of figures

Figure 1: Conceptual model ... 20

Figure 2: Main paths ... 31

(8)

1. Introduction

As the world today is changing ever so fast the general view of business has changed as well under the influence of globalisation of markets (Gulati, 1995). Furthermore, economists and management theorists are more concerned with the contractual or governance structures used in alliances (Gulati 1995). Additionally, changes in the economic environment influence the interfirm relations, therefore also affected are the relationships between buyer and supplier. In the literature we find evidence that the relationship between buyer and supplier is of central importance for the management in different firms, also that there is significant difference in the perceptions between buyers and suppliers, with regard to the performance of the relationship (Harland, 1996). The older studies are base for the work presented in some of the more recent studies, which build on and also prove that both buyers and suppliers perceive differently their relationships (Ambrose et al. 2010). Additional finding is that the “perceptual differences might influence the way attributes are associated” (Oosterhuis et al. 2013). The insufficient research information toward the separate perceptions of buyer and supplier leads to the conclusion that there is open space for further exploration on the topic of buyer-supplier relations. Moreover, because of the differences concerning perceptions and associations we offer a more in debt research toward the possible variations in the buyer-supplier relationship. We focus on studying the viewpoints of the two partners independently and furthermore in order to expand our research, we investigate the differences, reflected by the short or long duration of the relationship.

Why conduct this study

(9)

Master Thesis 2015| Margarita Racheva |9 contribute for the generalisation of the findings, as in the database which will be used the surveyed companies are from three different industries. Moreover, in the buyer - supplier relationship there are firms with various duration of the relationship. Furthermore, the survey is filled by employees occupying different positions in the company. The provided data sample which will be discussed later in the methodology plan, has the needed information and allows to avoid the previously described limitations and add to the further research by generalising the findings.

Additional support for the purpose of the research can be found in the work of Oosterhuis et al. (2013), who state that no specific attention was paid to the consequence of the perception differences between the buyer and the supplier and therefore, it is unclear how these perceptions influence the buyer-supplier relationship. Moreover, that there are no conclusive answers to the question whether buyers and suppliers have the same perceptions of their relationship. As possibility for future research they suggest to focus on the perceptual differences and their relation to the performance, additionally to study the conditions under which the difference in perceptions is positively or negatively related to the relationship performance.

General aims of the study

The goal of this research is first to analyse the role that governance has in the interfirm relations. Moreover, it aims to determine the degree to which the two studied governance modes, namely contracting (formal agreements) and relational norms (informal agreements) matter in the relationship for both partners. In addition, the influence that the governance modes have on the absorptive capacity (ACAP) and ultimately on the explorative and exploitative learning performance is studied.

Second general aim for the study is related to the two different points of view - the perceptions of the buyer and the supplier, which have been found to differ in certain dimensions (Ambrose et al., 2010). To be more accurate in our results, we will investigate dyad relationships between firms, which will allow for a greater understanding of both perspectives in the relationship.

(10)

Master Thesis 2015| Margarita Racheva |10 relationship will give an answer and fill the gap in the existing literature, which does not provide sufficient information for the short relationship, as it focuses mainly on the long term relationships.

The research will use the answers provided by 166 dyadic buyer-supplier relationships to formulate the main findings and will try to provide further support for them by eliminating some limitations found in previous researches.

Research question

Based on the already discussed findings, the following research question can be derived:

► Are the perceptions and associations regarding governance, absorptive capacity and learning performance different for buyer and supplier and does the duration of the relationship have an effect on the viewpoints of the two partners?

Procedure

(11)

2. Theoretical framework

This research is focusing on the link between governance forms like contracting and relational norms and their relationship with absorptive capacity for buyers and suppliers. Additionally, the relationship between absorptive capacity and learning performance is being examined, therefore it is important to describe the main concept found in literature of each of those components before presenting the relations between them.

First, we will focus on the absorptive capacity, its two sub parts: potential and realised absorptive capacity and we will further examine their underlying capabilities. Secondly, we will present the governance modes on which the study will focus: relational norms and contracting. Thirdly, the research will study the relationship between the potential absorptive capacity and the explorative learning performance and the relationship between realised absorptive capacity and the exploitative learning performance. Finally, we will investigate the results of the buyer and supplier perspectives and associations, which is the main focus of the study. The research will determine whether significant differences exist between the viewpoints of the two parties and additionally if the duration of the relationship has an effect on their perceptions and associations.

2.1 ACAP

The absorptive capacity (ACAP) is linked to the possibility to use knowledge that is external to the firm in order to benefit from various capabilities, related to innovations. In general it represents the ability to recognise, assimilate and apply new information to commercial ends (Cohen et al. 1990). Absorptive capacity has been centre of attention for many researchers (Lane et al. 2006) and its importance has been seen in many different fields like technological and strategic management, as well as organisational economics and international business (Zahra and George, 2002). Due to the complexity of the topic of ACAP, Zahra and George (2002) present the ACAP as being composed of two sub parts which are the potential and the realised absorptive capacity, as follows PACAP and RACAP.

(12)

Master Thesis 2015| Margarita Racheva |12 Authors explain the difference between the two dimensions, stating that PACAP focuses on the efforts spend in identifying and acquiring new external knowledge and assimilating that knowledge, however does not guarantee the exploitation of this knowledge (Jansen et al. 2005, Zahra and George, 2002) as its main focus involves adding new resources and competences (March, 1991). On the other hand, RACAP encompasses the derivation of new insights and consequences, combining the existing and the new knowledge and implementing the transformed knowledge into the procedures (Jansen et al. 2005, Zahra and George, 2002). It is primary used for further refining the existing competences (March, 1991). Following, we present a brief explanation of each capability, which is connected to PACAP and RACAP. Acquisition represents the firm’s ability to recognise and acquire new knowledge that is external for the firm and is important for its operation (Zahra and George, 2002; Leal-Rodríguez and Roldán, 2013). Assimilation represents the processes in the firm that allow the analysis, understanding and interpretation of the information collected from the external sources (Szulanski, 1996; Kim 1997a, b). As external knowledge is especially specific in the context that it is used, it could create barriers for outsiders to replicate the knowledge (Szulanski, 1996). Transformation represents the capability of the firm to develop and adapt the work routines while combining the previous, the new and the assimilated knowledge and trough these activities new schemas of action are created which can generate new competencies for the firm (Zahra and George 2002).Exploitation represents the ability of firms to incorporate the newly gained knowledge into the work operations and this way to create new competencies or to leverage existing ones (Zahra and George 2002).

2.2 Governance

According to theories, like the transaction cost theory, the governance in interfirm relationships will be anticipated by the asset specificity or the extent of certain specific investments that are involved in the transaction (Ambrose et al. 2010). In this research, the focus falls on the different governance modes that can be used in the buyer-supplier relationship. This is due to the fact that in literature the governance mode concerning the interfirm relations is a defining feature (Williamson, 1996) and it has been used to explain the dependence in the relationship between the two partners.

(13)

Master Thesis 2015| Margarita Racheva |13 contracting has been discussed and defined by authors who study Transactional Cost Economics (TCE). These authors describe transaction costs as typically arising from concerns about opportunistic behaviour by one of the partners. Additionally, they are caused by the inclusion of negotiation costs, monitoring of the contractual performance and the enforcement of contractual promises (Joskow, 1985). Furthermore, in TCE different assets associated with relations are considered as a potential source of opportunism (Berger 2015). Opposed to the TCE is the Resource Based View (RBV). It has been described as a source of competitive advantage, based on interfirm knowledge-sharing routines, through which the partner firms generate relational rents.

Based on these findings we can determine that there is not a clear path in the choice of governance modes between two partner firms. Moreover, there are additional contradictions concerning the two points of view towards formal contracting and on the other hand the informal relational norms.

Some researchers like Faems et al. (2008) say that the contracts that concentrate mainly on safeguarding are less capable of creating behaviour towards cooperation than those, which rely more on the interdependence and coordination between the parties. This finding is also supported by authors like Weber and Mayer (2011), who state that parties tend to cooperate more and be more creative when the contracts are more flexible and are aimed at achieving the main goal for both firms. In the same line of reasoning, Lumineau and Malhotra (2011) also contribute with their finding that if the contract points out the importance of coordination, then both partners together resolve the possible problems.

(14)

2.3 Relational norms

In the academic literature relational norms are described as actions and behaviour that are considered to be appropriate in the relationship, representing the shared values of the partners. (Heide and John, 1992; Morgan and Hunt, 1994; Joshi and Arnold, 1998). Relational norms include solidarity, flexibility, bilateralism as well as continuance and can create confidence and cooperation in the interfirm exchanges (Poppo and Zenger, 2002). Higher levels of commitment are also expected in relationships where the relational norms are high (Gundlach, Achrol, and Mentzer, 1995; Joshi and Arnold, 1998). The importance of relational norms is presented via the relationships created between the members as they communicate and become more familiar with each other they create mutual values and trust, which makes them more open and willing to share specific and sensitive information (Sun and Anderson, 2010). Such trustworthy behaviour from both partners is likely to have a positive effect on PACAP as the general expectations for relational norms are for them to be based on flexibility, experimentation and taking risks (Berger, 2015) and are in line with the capabilities related to PACAP to acquire and assimilate new information. Based on that we can formulate our first hypothesis - H1 which is stated below.

(15)

Master Thesis 2015| Margarita Racheva |15 absorptive capacity relates to exploitative learning performance. Furthermore, based on previous research, learning has been approached as both organisational and interorganizational phenomenon in which both parties jointly share, interpret and integrate information (Selnes and Sallis, 2003). This process has been found to breed trust between the partner firms on one hand and to reduce the outcomes of relational learning on performance on the other. Authors refer to this phenomenon as the “hidden costs of high trust” (Selnes and Sallis, 2003). Moreover, based on the consequences of this effect, discussed by Selnes and Sallis (2003) and how it affects the learning performance of the relationship, we can assume that there will be a positive effect between relational norms and both PACAP and RACAP, however, it will diminish when the level of relational norms is high (Berger, 2015). Therefore, based on this we can derive two of our hypotheses.

Hypothesis 1: There will be a positive relationship between relational norms and PACAP. Hypothesis 2: There will be a positive relationship between relational norms and RACAP.

2.4 Contracting

(16)

Master Thesis 2015| Margarita Racheva |16 that members seek new solutions. Taking these findings as a base and the fact that PACAP is related to explorative learning as well as the capabilities of acquisition and assimilation, we can assume that contracting and PACAP will be negatively related. Therefore, we can derive our third hypothesis regarding PACAP, which is the following:

Hypothesis 3: There will be a negative relationship between contracting and PACAP.

What we further find in the work of Jansen et al. (2005) is that the formalization in the relations will increase the possibilities that the members will recognise opportunities for the transformation of the new external knowledge. Moreover, that it will help the different units to codify and benefit from the use of the best practices, making the new knowledge more accessible and easy to use, therefore transformation and exploitation will be enhanced by formalisation (Jansen et al., 2005). Additional findings show that there is a positive relationship between the extent to which procedures and rules are implied within an organisation and exploitative innovations (Jansen et al., 2006). Relying on this, we can conclude that contracting will have a positive link with transformation and exploitation and therefore with RACAP, as those are its underlying capabilities. This additional perspective towards formalization as part of contracting, leads to our fourth hypothesis – H4.

Hypothesis 4: There will be a positive relationship between contracting and RACAP.

2.5 Explorative Learning Performance

(17)

Master Thesis 2015| Margarita Racheva |17 and contracting and PACAP and additionally the expected positive relationship between PACAP and explorative learning performance, we assume that PACAP will be a necessary condition to transfer the assets of the two types of governance modes to the learning performance. Therefore, we expect that PACAP will mediate the direct relationship between relational norms and explorative learning performance and additionally the relation between contracting and explorative learning performance. Based on this we formulate two additional hypotheses – H6a and H6b.

Hypothesis 5: There will be a positive relationship between PACAP and explorative learning

performance.

Hypothesis 6a: PACAP will mediate the relationship between relational norms and

explorative learning performance.

Hypothesis 6b: PACAP will mediate the relationship between contracting and explorative

learning performance.

2.6 Exploitative Learning Performance

(18)

Master Thesis 2015| Margarita Racheva |18 performance. Based on this, we formulate two more hypotheses – H8a and H8b, which are related to RACAP.

Hypothesis 7: There will be a positive relationship between RACAP and exploitative learning

performance.

Hypothesis 8a: RACAP will mediate the relationship between relational norms and

exploitative learning performance.

Hypothesis 8b: RACAP will mediate the relationship between contracting and exploitative

learning performance.

2.7 Buyer - Supplier perspective

Following the work of previous researchers, we find that there is proof of the different perceptions of buyer and supplier regarding the relationship. In a study made by Shin et al. (2000), the researchers find that data taken from the buyer shows a positive link between the buyer relationship and the supplier performance. However other studies like the one from Prahinski and Benton (2004) don’t support such relationship. Such findings show that a difference exists in the perceptions of the participants in the relationship, also that hypotheses supported by data only from the buyer or the supplier might differ from each other and therefore, question the validity and reliability of researches that use data from only one of the participants in the relationship (Oosterhuis et al. 2013). Authors like Oosterhuis et al. (2013) state that these perceptions diverge most of the time and that they may influence the reliability of the measurements and also the association of different attributes.

(19)

Master Thesis 2015| Margarita Racheva |19 feeling of belonging to the organisation of either buyer or supplier and additionally they may create biases for one of the organisations resulting in more favourable perceptions towards certain aspects (Oosterhuis et al. 2013).

Derived from the findings discussed earlier in our paper, concerning the different relations, unique experiences and knowledge basis between the two firms and how these factors create information asymmetry, we expect to find differences between the viewpoints of the buyer and supplier towards the main constructs studied in this research.

Based on these findings we can continue and study the differences between buyer and supplier, however in order to try and obtain a deeper understanding in the interfirm relationships this study also focuses on the duration of the relationship. Studying the buyer - supplier relationship, there are authors who focus on the length of the relationship. Some of the findings suggest that supplier firms engaging in long-term relationships are more capable of achieving higher profitability (Kalwani and Narayandas, 1995). In later works such conclusions are backed up by other studies, like the one of Dyer and Chu (2000), who have also concentrated on the positive relationship between the levels of trust and the time the two companies have been working together. Additionally, their findings explain that the longer duration since the first buyer – supplier transaction, the higher the supplier’s trust in the buyer will be. Therefore, we expect to find stronger positive effects in the buyer-supplier relationships that have existed for a longer period of time.

(20)

Figure 1: Conceptual model

RACAP Relational Norms

Contracting

PACAP Explorative Learning

(21)

3. Methodology plan

“There are two sides of every coin” - Zhang Guobao

Many research papers base their findings only on one of the participants in the relationship. Focusing on either buyer or supplier creates significant constraints for the analysis and poses a question to the validity of the findings, as both sides in a business relationship see differently the real situation. To avoid limitations caused by the data received only from one of the partners in the relationship, the study is based on dyadic data collected from both supplier and buyer. Following the example of other authors like Selnes and Sallis (2003), we examine the relationship between the two parties and not on the individual unit. Reason for this is the possibility of measurement errors when information is used from only one party of the relationship, which can be eliminated by focusing on the dyadic relationships (Selnes and Sallis, 2003). Many authors have researched the effect that such data has on the results of the studies and many of them point out also the importance of dyadic business relationships (Anderson et al., 1994). In such case, studying the perceptions of both parties is necessary in order to derive valid statements and findings. Moreover, the information from both sides eliminates the possible constraints, concerning the unsatisfactory amount of information received from only one of the firms. Therefore we can say that the dyadic information from both partners gives an advantage to the present research.

3.1 Data collection and Questionnaire design

The data used for the present research is acquired by the study of Berger (2015). The questionnaire was made for a survey on Interfirm Absorptive Capacity in 2010 and was initiated by the faculty of Economics and Business in the University of Groningen. It includes 166 dyadic relationships between buyer and supplier. The parts of the questionnaire used for this study can be seen in Appendix A.

(22)

Filling the survey were key informants from both companies. They answered questions divided in 5 main categories. Apart from the first one, all other categories contain questions for which the answers are presented in a 7-point Likert scale in which 1 represents “strongly disagree” and 7 represents “strongly agree”. There are some minor differences in the questions that aim to measure explorative learning performance, where the two ends of the scale are representing “to no extent” and “to a great extent”. For the first sector of questions - ”General questions” the informants had to fill in more personal information about their position within the firm, like the number of years they were occupying their current position and the % of time they spent on activities related to the supplier-buyer relationship. This section also includes the more important but also sensible information, like sales volumes or spendings, which many companies are reluctant to share.

3.2 Measurements

According to the information provided along with the questionnaire from the previous study (Berger, 2015), a pre-test was made with several buyers and suppliers in order to ensure the clarity of the questionnaire by making the necessary adjustments on it.

With regard to the measurements, we follow the initial study that provides the questionnaire and the dataset where the measures for contracting were taken from different papers including Buvik and Reve (2002) and more recent ones like Jansen et al. (2006). Concerning the measures for the other type of governance, namely relational norms, they were retrieved using the work of Heide and John (1992).

As it has already been mentioned, the two dimensions of absorptive capacity both refer to the learning capabilities, namely PACAP is linked to explorative learning and RACAP is linked to exploitative learning. The measures used for the explorative learning performance are based on the work of Lane et al. (2001), as for exploitative learning performance the measure is used based on the work of Selnes and Sallis (2003).

(23)

In our research, two types of constructs are used, namely first-order and second-order ones. In the case of first-order constructs, they are observing different variables as indicators and in the case of second-order they are based on other first-order constructs (Berger, 2015). Additionally, we take into account the difference between reflective and formative scales. Reflective scales are characterised with strong correlations between the components (Berger, 2015), as in the case of formative scales, they are composed of indicators measuring various aspects of the construct (Hair at al., 2011). The research is composed of three second-order constructs, namely PACAP, RACAP and Relational norms, where all other constructs are first-order ones.

3.3 Statistical procedures

One of the main concepts studied in this work is related to the mediating variables in the relationships. As it is being defined in previous scientific papers, a certain variable may have the function of a mediator when it “accounts for the relation between the predictor and the criterion” (Baron and Kenny, 1986). To determine the existence of mediation, regression analysis is often used, however this technique assumes that there is no measurement error in the mediator which entails that the effect of the mediator on the dependent variable cannot be controlled when the effect of the independent variable on the dependent variable is evaluated (Berger, 2015) In order to avoid such errors, authors suggest the use of structural modelling methods (Baron and Kenny, 1986). These findings are stimulus for the decision to choose a structural equation model (SEM) for the research.

(24)

Master Thesis 2015| Margarita Racheva |24 Berger 2015). In the case of our research the split data provides 166 samples for both buyer and supplier, further investigating the effects of duration on the relationship and splitting the data once more the research will have approximately 84 subjects as sample size, which is sufficient for the required amount for the sample.

3.4 Research design

The goal of this research is to analyse the influence of factors like relational norms and contracting on the Potential and Realised Absorptive capacity (PACAP and RACAP), while focusing on both perspectives from the buyer and the supplier and considering the duration of the interfirm relationship. Additionally, we analyse the links between PACAP and the explorative learning performance and RACAP and the exploitative learning performance. Moreover, as we hypothesise that PACAP and RACAP will be necessary conditions in order to transform the assets of the two types of governance into learning performance, we conduct mediation analysis to determine whether or not a mediation effect exists between the constructs.

In order to investigate the validity of the hypotheses mentioned in the previous part, we use certain information from the data collected via the questionnaire (Appendix A). Moreover, we split the data in order to be able to distinguish between the conditions concerning the buyer and supplier and furthermore between the short and long duration of the relationship. Presented below is a detailed explanation of each of the variables in the model and its construct.

Independent Variables

(25)

Mediators

The model suggests two moderators - PACAP and RACAP, which both are dimensions of ACAP. We expect that PACAP will be mediating the relationship between relational norms as an independent variable and explorative learning performance as a dependent variable, also the relationship between contracting and explorative learning performance. Following the same logic, we expect that RACAP will be mediating the relationship between contracting and exploitative learning performance as well as the relationship between relational norms and exploitative learning performance. As PACAP and RACAP both have two underlying capabilities we use the questions related to those capabilities in order to study the effects of PACAP and RACAP. Therefore, we use the questions related to the ability of the relationship to recognise and acquire new external knowledge (Appendix 1, Q.1) and the questions related to drawing sense out of the newly acquired external knowledge (Appendix 1, Q.2) to formulate the variable for PACAP. Additionally, we follow the same procedure in creating the RACAP variable, in which we include the information regarding the combining of old and new knowledge, the process of sharing and transferring external knowledge to the relationship and the information about commercializing new external knowledge (Appendix 1, Q. 3; 4; 5).

Dependent Variables

In the model there are two dependent variables representing the learning performance: explorative learning performance which is linked to PACAP and exploitative learning performance which is linked to RACAP. In order to create the variable for explorative learning performance we use the questions regarding the extent of expertise (Appendix 1, Q. 8a). Likewise, for exploitative learning performance we use the questions related to the performance of the relationship (Appendix 1, Q. 9a).

4. Analysis and Results

(26)

4.1 Measurement Model

In order to measure the model for consistency, the total data is being tested, as well as the data from both parties independently, meaning that we use the already separated data in SPSS and import the two databases – for buyers only and for suppliers only. In our study, as the initial database consists of 332 respondents, when the data is separated we have 166 subjects in each case.

While assessing the data we must distinguish between the two types of scales that are part of the construct of the variables and which exist in our model. These are formative and reflective scales. Such distinction has to be made because of the different way in which tests are performed for the formative and reflective scales. In the case of reflective indicators, they are represented as functions of the latent construct and therefore changes in the indicators are reflected also as changes in the construct itself (Hair et al., 2011). On the other hand, formative indicators are assumed to cause the latent constructs, therefore any changes made to the indicators can cause value changes in the latent construct (Hair et al., 2011).

Therefore we examine the data by assessing separately the validity of reflective and formative scales. For reflective scales, we study the values of the loadings, Cronbach’s alpha and the Composite Reliability as presented in Table 1 below. The measures used to determine the validity of the results suggest that the indicator loadings should be equal or greater than 0.5 (Hair et al. 2010), the requirements for the p-values are for them to be lower than 0.05 in order to be valid. Following the example of Berger (2015), in checking for reliability in PLS both Cronbach’s alpha and Composite reliability are studied. The two analysis provide the same information, however for PLS it is better to consider both of them as Cronbach’s alpha assesses the model assuming that all indicators add equally distributed to the overall reliability (Henseler, Ringle and Sinkovics, 2009), whereas composite reliability estimates the true factor loadings of the indicators (Götz, Liehr – Gobbers, and Krafft, 2010). For the two coefficients to show sufficient reliability, they must have values equal or greater than 0.70 (Nunnally and Bernstein, 1994), additionally if the value is less than 0.60 this indicates that there is lack of reliability (Berger, 2015).

(27)

(28)

Master Thesis 2015| Margarita Racheva |28 For the Formative scales, we take into account the values of the weights, the variance inflation factors (VIF) and the p-values after performing a bootstrap. The results are shown in Table 2.

Subsequently, the measures used for determining the validity of the values is presented. For the p-values to be valid they have to satisfy the condition to be 0.05 or lower. In the case of the variance inflation factors (VIF), they have to have values lower than the cut-off point of 3.3 (Petter, Straub and Rai, 2007; Cenfetelli and Bassellier, 2009).

Based on the analysis of the data 4 variables were removed from the data. The variables 1.2 for acquisition and 3.3 for transformation had an insufficient weights, as follows of 0.058 and 0.060 and were removed. Moreover, variables 3.2 and 4.1 for transformation were also deleted because of inadequate p-values, which were as follows 0.096 and 0.058. Because these four variables are related to formative scales we carefully investigate whether it is appropriate to remove them from the model. Based on the content of the questions corresponding to those variables we can conclude that their deletion will not have an effect on the general content of the formative constructs. To avoid inconsistency in the results that our research aims for, namely the comparison between buyer and supplier we remove these 4 variables from both buyer only and supplier only datasets. This way we ensure that the results for each aspect received from both parties will be comparable. Below Table 2 is presented, including the combined data analysis for buyer and supplier without the already mentioned deleted questions.

Table 2: Convergent validity of formative scales.

Constructs

/indicators:

Buyer and Supplier Indicator

weights: p-value: VIF:

Acquisition Var.1.1 .306 <.001 1.279 Var.1.3 .308 <.001 1.190 Var.1.4 .385 <.001 1.413 Var.1.5 .379 <.001 1.440 Transformation Var.3.1 .332 <.001 1.635 Var.3.4 .205 <.001 1.206 Var.4.2 .299 <.001 2.549 Var.4.3 .294 <.001 2.218 Var.4.4 .235 <.001 2.267

Explorative Learn. Perf.

(29)

Var.8.2 .264 <.001 1.936

Var.8.3 .270 <.001 1.996

Var.8.4 .241 <.001 1.873

Var.8.5 .248 <.001 1.958

Exploitive Learn. Perf.

Var.9.1 .171 <.001 1.454 Var.9.2 .233 <.001 1.680 Var.9.3 .227 <.001 1.370 Var.9.4 .192 <.001 1.400 Var.9.5 .250 <.001 1.693 Var.9.6 .251 <.001 1.866 Var.9.7 .237 <.001 1.435 PACAP (2nd order construct) Acquisition .456 <.001 1.672 Assimilation .647 <.001 1.927 RACAP (2nd order construct) Transformation .620 <.001 2.106 Exploitation .457 <.001 1.920

From the results we can see that all the indicators have p-values lower than the required 0.05. Additionally, all VIF values are lower than the cut-off point of 3.3 (Cenfetelli and Bassellier, 2009). Therefore, we can conclude that these results show the valid measures for the formative scales.

As a continuation of the analysis we study the discriminant validity by comparing the square roots of the average variance extracted (AVE) with the correlation coefficients. This type of analysis is used for reflective and formative scales (Berger, 2015). In the following Table 3a, and 3b, are presented the discriminant validity results for the dataset, where the first and second order variables are presented separately. The two types of data are presented together as the squared roots of the average variance extracted (AVE) are presented along the diagonal and the other values represent the correlation coefficients.

Table 3a: Squared roots of the average variance extracted (AVE) and correlation matrix. First

order constructs.

1 2 3 4 5 6 7 8 9 10

1. Acquisition 0.521

2. Assimilation 0.631 0.541

(30)

Master Thesis 2015| Margarita Racheva |30 4. Expliotative LP 0.395 0.480 0.114 0.405 5. Exploitation 0.533 0.636 0.105 0.495 0.516 6. Explorative LP 0.263 0.308 0.142 0.575 0.301 0.593 7. Flexibility 0.187 0.348 -0.010 0.285 0.348 0.024 0.648 8. Info.Exchange 0.313 0.424 -0.012 0.306 0.382 0.150 0.480 0.595 9. Solidarity 0.270 0.400 0.035 0.315 0.388 0.169 0.516 0.661 0.748 10. Transformation 0.538 0.653 0.095 0.462 0.676 0.278 0.327 0.508 0.434 0.532

Table 3b: Squared roots of the average variance extracted (AVE) and correlation matrix.

Second order constructs.

1 2 3

1. PACAP 0.813

2. RACAP 0.724 0.841

3. Relational Norms 0.443 0.525 0.702

Based on the results from the analysis, where the diagonal AVE scores are higher than the values in the corresponding columns or rows, we can conclude that the model as an overall is strong.

4.2 Structural model analysis

This section will present the model of the research, where the conceptual model will be studied in two steps. One will relate to the direct effects and another to the possible indirect or mediating effects. The direct effects concern whether or not direct links exist between the variables and the indirect effects will determine whether some of the variables mediate the indirect relationships.

Guided by the main focus of the research, which is to study the difference between buyer and supplier and how they perceive and associate the interfirm relationship, the first step taken for analysing the data is to separate the dataset in two, in order to facilitate the exploitation of information independently for buyers and suppliers. Furthermore, our study investigates whether there are differences between the short and long term relationships for both buyer and supplier. This requires the separation of the data for a second time after the initial split which determined the validity of the information. The variable we use for the separation is the duration of the relationship, for which we use the mean in order to be able to separate the data in such way that there will be enough subjects in each condition.

(31)

Master Thesis 2015| Margarita Racheva |31 performance, where PACAP is a mediator (Fig. 2.1). Second, contracting is linked with PACAP and ultimately to explorative learning performance, where PACAP is a mediator (Fig 2.2). Third, relational norms have an effect on RACAP and ultimately on exploitative learning performance, where RACAP is a mediator (Fig 2.3). Fourth, contracting is linked to RACAP and ultimately to exploitative learning performance, where RACAP is a mediator (Fig 2.4). The four paths are visually presented below in Figure 2.

Figure 2: Main paths

RACAP

Relational Norms

Contracting

PACAP Explorative Learning _Performance

Exploitative Learning Performance H1 H2 H3 H4 H5 H7 H6a,b H8a,b RACAP Relational Norms Contracting

PACAP Explorative Learning Performance Exploitative Learning Performance H1 H2 H3 H4 H5 H7 H6a,b H8a,b Fig.2.1 Fig.2.2 RACAP Relational Norms Contracting

PACAP Explorative Learning _Performance

Exploitative Learning Performance H1 H2 H3 H4 H5 H7 H6a,b H8a,b RACAP Relational Norms Contracting

(32)

4.2.1 Direct effects

In the conceptual model several direct effects are being presented. Therefore in the following Table 4 the results for those direct effects can be seen. Presented are the path coefficients and the p-values which determine the validity of the relationship.

The table shows the relation between each independent variable and its effect on the dependent one. Moreover, the data suggests 4 general cases by using the split data into buyer short term and long term relationship and supplier short and long term relationship.

Table 4: Standardized path coefficients Independent

Buyer Supplier

short relationship long relationship short relationship long relationship dependent:

Beta: p-value: Beta: p-value: Beta: p-value: Beta: p-value: Relational Norms PACAP .570 <.001 .558 <.001 .471 <.001 .188 .124 RACAP .644 <.001 .571 <.001 .636 <.001 .216 .135 Contracting PACAP .146 .101 .316 <.001 .163 .054 -.057 .628 RACAP .061 .464 .287 <.001 .139 .103 .055 .689 PACAP Explorative LP .220 .043 .493 <.001 .316 .005 .227 .035 RACAP Exploitative LP .407 <.001 .760 <.001 .476 <.001 .437 <.001 From Table 4 we see that there are differences in the values, however to be able to draw conclusions we must understand whether the differences in the path coefficients are significant or not. To investigate the difference between the conditions we perform a Multi-group analysis (MGA) in PLS using the different Multi-groups in our study, namely the two Multi-groups for buyer with short and long duration of the relationship and the two groups for the supplier, again separated by the duration of the relationship to long and short.

(33)

Master Thesis 2015| Margarita Racheva |33 differences is reflected in the larger values that one of the two groups in the Multi-group analysis has. If the p-values show significant difference with a value of 0.05 or lower that means that the results for the 2nd group are almost always larger than the 1st group, relatively the reversed results apply if the p-values show significant difference with value of 0.95 or higher, then the results of the 1st group are often larger than the ones of the 2nd group. (Henseler, 2007; Sarstedt, Henseler, and Ringle, 2011).

In Table 5a is shown a Multi-group analysis comparing the 2 possible conditions for the duration of the relationship for each partner. We find that, for the short and long term relationship of the buyer there are significant differences in the effect of contracting on RACAP, the effect of PACAP on explorative learning performance and the effect of RACAP on exploitative learning performance. As the significant difference is with values of 0.95 and higher we can conclude that the results for group 1 (in our case buyer short duration) are often larger than those of group 2 (buyers long duration). The MGA for the supplier shows 2 significant differences, which apply for the effect of relational norms on PACAP and RACAP. In this case the significant difference has values of 0.05 or lower, therefore in both cases the results of group 2 (buyer long duration) are larger than the results of group 1 (buyers short duration). For the effect of contracting on PACAP no significant differences were found in both buyer and supplier analysis.

Table 5a: Multi group analysis results, buyer - supplier data.

Independent Multi-group analysis

Buyer short/ long duration Supplier short/ long duration dependent: _{Path Coefficients}

differences p-value: Path Coefficients differences p-value: Relational Norms PACAP 0.011 0.466 0.283 0.058 RACAP 0.073 0.303 0.421 0.006 Contracting PACAP 0.170 0.906 0.220 0.068 RACAP 0.226 0.977 0.084 0.307 PACAP Explorative LP 0.256 0.985 0.208 0.113 RACAP Exploitative LP 0.244 0.951 0.114 0.772

(34)

Master Thesis 2015| Margarita Racheva |34 that none of the p-values in the first part of the table is significant according to the criteria we discussed previously. Therefore, we can conclude that there are no differences in the way partners with short relationships perceive the interfirm relations. However, in the case of long relationships four out of the six effects between the constructs show significant difference. There are only two relations, namely the effect of contracting on RACAP and the effect of RACAP on exploitative learning performance that do not show significant difference. Furthermore, we have to mention that the effect of contracting on RACAP shows value very close to the significant cut-off point. As the significant values are 0.05 or lower, we conclude that the results of group 2 (suppliers with long duration of the relationship) are larger than the results of group 1 (buyers with long duration of the relationship).

Table 5b: Multi group analysis results, short - long duration. Independent

Multi-group analysis

Buyer short/ Supplier short duration Buyer long/ Supplier long duration dependent: _{Path Coefficients}

(35)

4.4.3 Indirect / mediation effects

To better explain the appearance of an indirect or mediation effect in our study, we will shortly explain the concept of mediation. As it has been already mentioned in the methodology section, one variable can be considered as having a mediating effect when it “accounts for the reaction between the predictor and the criterion” (Baron and Kenny, 1986). To present this effect visually we use the work of Preacher and Hayes (2004), who illustrate it as shown in Figure 3, provided below.

Figure 3: Direct effect; Mediation effect

(36)

Master Thesis 2015| Margarita Racheva |36 In our research we test for mediation as we study each path of the conceptual model independently for each of the four groups, buyer short and long duration of the relationship and supplier short and long duration of the relationship. Using PLS, we first check the direct effect between the independent and dependent variable, next we add the moderator and compare the differences in the values. Additionally, we take in consideration the Beta, the Standard Error and the T-Statistics. The first two we use to get results for Sobel test, which will determine if the indirect effect is significant (Sobel, 1982). We choose for the Sobel test because it provides a more direct test on the mediation effects (Preacher and Hayes, 2004). It compares the strength of the indirect effect between X and Y, which is represented by the two effects of X of M and of M on Y or the ab path with the to the point null hypothesis, that equals to zero (Preacher and Hayes, 2004). After we have obtained the values for the Beta and the Standard Error in PLS, we use the software “Sobel test calculator for the significance of mediation” (Soper, 2015) to determine the values of the Sobel test and the Two-tailed probability. In order for the values to be significant and therefore for mediation to exist the Sobel test must have absolute values of 1.96 or higher, and the Two-tailed probability has to be 0.05 or lower (Preacher and Hayes, 2004).

The results from the analysis are presented in Tables 6a, b and c below. The first table (a) represents the results on the mediation analysis for the first separation of the data between buyer and supplier, additionally tables 6b and 6c study the possible mediation effects after the second split of the data made according to the duration of the relationship.

(37)

(38)

IV-M M-DV IV-M M-DV 1 RN 0.128 -0.070 0.523 0.379 0.067 0.086 0.927 3.837 <.001 ✓ 2 Contr 0.203 0.160 0.134 0.321 0.088 0.077 1.972 1.430 0.152 x 3 RN 0.366 0.040 0.594 0.549 0.072 0.088 0.438 4.976 <.001 ✓ 4 Contr 0.094 0.031 0.072 0.878 0.088 0.021 0.841 0.818 0.413 x 1 RN 0.149 0.052 0.366 0.267 0.099 0.077 0.786 2.529 0.011 ✓ 2 Contr -0.001 -0.035 0.117 0.290 0.072 0.074 0.452 1.501 0.133 x 3 RN 0.359 0.180 0.468 0.382 0.089 0.079 1.936 3.559 <.001 ✓ 4 Contr 0.112 -0.048 0.178 0.895 0.076 0.019 1.172 2.339 0.019 ✓ IV-M M-DV IV-M M-DV 1 RN 0.133 0.022 0.535 0.208 0.097 0.146 0.183 1.379 0.167 x 2 Contr 0.021 0.019 0.011 0.219 0.121 0.116 0.166 0.090 0.927 x 3 RN 0.527 -0.032 0.630 0.888 0.109 0.060 0.425 5.383 <.001 ✓ 4 Contr -0.062 0.018 0.092 0.869 0.109 0.030 0.326 0.843 0.398 x 1 RN 0.160 -0.148 0.538 0.573 0.089 0.104 1.390 4.072 <.001 ✓ 2 Contr 0.457 0.346 0.280 0.397 0.113 0.082 2.832 2.205 0.027 ✓ 3 RN 0.412 -0.120 0.552 0.963 0.104 0.051 1.487 5.109 <.001 ✓ 4 Contr 0.257 0.035 0.249 0.888 0.102 0.031 0.676 2.432 0.015 ✓ IV-M M-DV IV-M M-DV 1 RN 0.230 0.097 0.495 0.268 0.126 0.124 0.782 1.893 0.058 x 2 Contr 0.008 -0.070 0.233 0.332 0.096 0.118 0.655 1.837 0.066 x 3 RN 0.565 -0.062 0.657 0.954 0.099 0.053 0.863 6.226 <.001 ✓ 4 Contr 0.142 -0.076 0.234 0.931 0.093 0.022 1.579 2.511 0.012 ✓ 1 RN 0.073 0.036 0.168 0.221 0.123 0.106 0.394 1.142 0.253 x 2 Contr 0.036 0.034 0.008 0.227 0.115 0.109 0.286 0.069 0.944 x 3 RN 0.182 -0.017 0.235 0.846 0.136 0.041 0.240 1.721 0.085 x 4 Contr 0.105 -0.005 0.131 0.843 0.123 0.032 0.064 1.064 0.287 x Result B u ye r PACAP Explor RACAP Exploit S u p p li er IV Mediator DV Direct No Med. Direct With Med.

Beta Stand.Error T-statistics of

direct effect PACAP Explor RACAP Exploit IV Mediator DV Sobel test statistic Two tailed probability Two tailed probability Result B u ye r Short PACAP Explor RACAP Exploit Long PACAP Direct No Med. Direct With Med.

direct effect

Sobel test statistic

direct effect Sobel test statistic Explor RACAP Exploit IV Mediator DV № № Explor RACAP Exploit Duration Duration № Two tailed probability Result S u p p li er Short PACAP Explor RACAP Exploit Long PACAP Direct No Med. Direct With Med.

Table 6: Mediation effects a

b )

(39)

5. Discussion

The discussion section will present the main findings from the research and also suggest possible implications. Furthermore, some limitations affecting our study and some additional ideas for future research will be presented at the end. In Table 7 are presented the hypotheses that are being researched in this study and whether they are supported or not according to the obtained results. The supported hypotheses are marked with “✓”, the rejected ones with “x” and for the ones that the study found no significance with “-“.

Table 7: Hypotheses

Hypotheses Buyer Supplier

short long short long

H1: There will be a positive relationship between

relational norms and PACAP. ✓ ✓ ✓ -

relational norms and RACAP. ✓ ✓ ✓ -

H3: There will be a negative relationship between

contracting and PACAP. - X X -

contracting and RACAP. - ✓ - -

PACAP and explorative learning performance. ✓ ✓ ✓ ✓

H6a: PACAP will mediate the relationship between

relational norms and explorative learning performance. X ✓ X X

H6b: PACAP will mediate the relationship between

contracting and explorative learning performance. X ✓ X X

RACAP and exploitative learning performance. ✓ ✓ ✓ ✓

H8a: RACAP will mediate the relationship between

relational norms and exploitative learning performance. ✓ ✓ ✓ X

H8b: RACAP will mediate the relationship between

contracting and exploitative learning performance. X ✓ ✓ X

(40)

Master Thesis 2015| Margarita Racheva |40 dimensions of ACAP. However, H1 and H2 have proven to be supported for buyers with both short and long duration of the relationship and also for suppliers with short duration of the relationship. This provides further support for the findings of authors like Selnes and Salis (2003) in the statement that both ACAP dimensions are influenced the same way when relational norms are used as governance mode. In the case of contracting on the two ACAP dimensions our research could not find all the necessary significant information in order to confirm all hypotheses for the 4 cases. The hypothesized negative effect of contracting on PACAP (H3) has not been supported for the buyers engaged in long relationships and the suppliers who have short relationships. Moreover, a positive effect has been found. However, the values showing significance for the short term supplier relationship are not showing very strong significant values (Table 4), which leaves us with only one case out of the four groups of subjects which is unsatisfactory for a clear statement, regarding the negative effects of contracting on PACAP. Combined with the contradictory findings in the work of Jansen (2005, 2006), this brings us to the conclusion that further research is needed with respect to the effect of contracting on PACAP. Similarly, for H4 the study again does not have sufficient support in the significance of the results, however in the case of buyers with long relationship the hypothesis is confirmed, supporting the findings of Jansen (2005, 2006).

Hypotheses 5 and 7 are supported for all four cases, which confirms the expected positive relationship between PACAP and explorative learning performance and RACAP and exploitative learning performance. This finding is in line with the early study of Zahra and George (2002) that associates PACAP with explorative learning performance and RACAP with exploitative learning performance.

The outcomes for hypotheses H6a and H6b show contradicting results. If we use the data regarding only the buyer and supplier data (Table 6a), without taking into account the duration of the relationship, PACAP is mediating the relationship between relational norms and explorative learning performance in both cases, which is supporting H6a. However, there is no mediation effect of PACAP on the relationship between contracting and explorative learning performance, which rejects H6b. Furthermore, if we look at the data with the separation by relationship duration, only for the buyers with long relationships a mediation effect exists.

(41)

Master Thesis 2015| Margarita Racheva |41 performance (Table 6a). Interesting to find is that when we look at the twice separated data the moderation effect exists in almost all cases. Exception is made by the suppliers with long duration of the relationship and in the case of buyers with short relationship for H8b.

General finding from the 4 hypotheses with mediation effects is that with the split of the data by the time involvement in the relationship in all cases for suppliers with long relationships there is no mediation for both PACAP and RACAP and mediation exists for all cases for buyers with long involvement in the relationship.

Based on the results in Table 5a and the hypotheses results we find that the effect of relational norms as a governance mode is perceived differently from buyers and suppliers. For buyers there is no significant difference between the perceptions of the short and long involvement, however in the case of buyers the difference is significant. Investigating the results for contracting, there are no significant differences for both parties regarding contracting and PACAP. However, for realised absorptive capacity in the case of buyers the differences are supported as these differences are significant for one of ACAPs dimensions but not the other. Furthermore, we find that there are differences with regard the effect between the two ACAP dimensions and learning performance and in the case of buyers they do have different perceptions and associations, however, this does not relate for suppliers.

To summarise, the results show that buyers and suppliers generally differ in their perceptions towards governance, absorptive capacity and learning performance. The only effect on which both parties have similar understandings is regarding the effect of contracting on potential absorptive capacity. This can be explained by the finding of Oosterhuis (2013) related to the access to knowledge, which explains that buyer and supplier can perceive the relationship differently if they possess different knowledge. However, in the case of contracting the two parties in the relationship have the same information, which explains our findings.

Based on the findings in Table 5b, we can conclude that there are no differences in the way partners with short relationships perceive the interfirm relations. However, in the case of long relationships in most cases significant differences exist. There are only two relations: the effect of contracting on RACAP and the effect of RACAP on exploitative learning performance that do not indicate significant differences.

(42)

Master Thesis 2015| Margarita Racheva |42 and 4 there are insufficient results for a clear statement to be drawn. Finally, for the hypotheses regarding the mediation effects, namely hypotheses 6a, 6b, 8a and 8b, there are contradictory findings. Considering PACAP as a mediator, in most cases the mediation is rejected, apart for the case of buyers with long duration of the relationship. Examining RACAP as a mediator in most of the cases the mediation hypotheses are supported. Exception is made by the suppliers with long relationships for both hypotheses 8a and 8b and the buyers with short relations for hypothesis 8b.

5.1 Implications

Based on the findings in this research we can present several possible implications. First, the research considers two main governance modes which are often used in interfirm relations. Gaining more knowledge on the viewpoints of the two parties in the firm can be beneficial for future research and can have significant value for the management of a certain firm. Furthermore, it could be helpful that the study provides insights related to the lifecycle phases, that each buyer and supplier encounter in the interfirm relationship in the beginning of the cooperation and at a more mature stage.

Second, the research considers both underlying dimensions of ACAP, namely PACAP and RACAP. Therefore, as there is a limited number of studies that investigate these dimensions separately and provide individual insights, the findings of this research are valuable addition to the literature. Moreover, this study creates better understanding of the general notion of ACAP and clearly separates the capabilities that refer to PACAP and RACAP.

(43)

5.2 Limitations and Future research

In this section both limitations and possibilities for future research will be discussed. Starting with the limitations, we will subsequently relate these limitations to the possibilities for future research.

The first limitation that can be suggested is concerning the data received from the respondents. The information from the dataset was used all together, however the questionnaire was distributed among companies from different industries. Because of that, no clear statement can be made for the buyer – supplier relationships in specific industries. This implies the possibility for future research projects that will gather more dyadic data on individual industries and therefore, it will be possible to make clear conclusions on the interfirm relations in certain sectors.

Second limitation to the study is that in order to research the buyer and supplier based on the duration of the relationship the initial data has to be split twice. Even with our study corresponding to the requirements for the analysis, this separation of the data reduces significantly the number of subjects per condition. Therefore, as a suggestion for future research the collection of larger number of data can be recommended when the study involves dyad relationships.

6. Conclusion

(44)

7. References:

Ambrose, E., Marshall, D, and Lynch, D. 2010. Buyer supplier perspectives on supply chain relationships. International Journal of Operations and Production Management, 30(12): 1269-1290.

Anderson, J. C., Håkansson, H., Johanson, J. 1994. Dyadic business relationships within a business network Context. Journal of Marketing, 58(4), 1-15.

Baron, R.M. and Kenny, D. A. 1986. The moderator – mediator variable distinction in social psychological research. Conceptual, strategic and statistical considerations. Journal of

personality and Social Psychology, 51 (6): 1173 – 1182.

Berger, J. 2015. Essays on the Governance of Buyer-Supplier Relationships. Groningen: University of Groningen.

Buvik, A., and Reve, T. 2002. Inter-firm governance and structural power in industrial relationships: The moderating effect of bargaining power on the contractual safeguarding of specific assets. Scandinavian Journal of Management, 18: 261 – 284.

Carr, A.S., Pearson, J.N., 1999. Strategically managed buyer–seller relationships and performance outcomes. Journal of Operations Management 17, 497–519.

Cenfetelli, R., and Bassellier, G. 2009. Interpretation of formative measurement in information system research. MIS Quarterly, 33 (4), 689 – 708.

Chin, W.W. 1998. The partial least square approach to structural equation modelling. In G. A. Marcoulides (Ed.) Modern methods of business research: 295 – 336. Mahwah, NJ: Laurence Erlbaum Associates.

Cohen, W.M., and Levinthal, D.A. 1990. Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35: 128-152.

Buyer and Supplier perspective in interfirm relations: The role of governance in the buyer - supplier relationship and its influence on ACAP and ultimately on learning performance

Buyer and Supplier perspective in interfirm relations:

The role of governance in the buyer - supplier relationship

and its influence on ACAP and ultimately on learning

performance

Buyer and Supplier perspective in interfirm relations:

The role of governance in the buyer - supplier relationship

and its influence on ACAP and ultimately on learning

performance

Master Thesis, MSc Marketing Management,

University of Groningen, Faculty of Economics and Business,

Department of Marketing

22 June 2015

Margarita Georgieva Racheva

Student number: S2699923

Address: Oude Riet 40, 9746 PL Groningen

Tel: + 31 (0)616759059

E-mail: margarita.racheva21@gmail.com

Abstract

Preface

Contents

List of Tables

List of figures

1. Introduction

2. Theoretical framework

3. Methodology plan

4. Analysis and Results

5. Discussion

6. Conclusion

7. References: