Supply Chain Resilience, learning from disruptions:

(1)

Supply Chain Resilience, learning from disruptions: A multiple case study in water supply industry.

Supply Chain Resilience, learning from

disruptions:

A multiple case study in the water supply industry.

Master thesis MSc. Supply Chain Management

University of Groningen, Faculty of Economics and Business P.O. Box 800, 9700 AZ Groningen

June 22, 2015 Cor Jonker Student number: 2234041 Petrus Campersingel 31-36 9713AD Groningen Tel: +31(0)611649357 c.jonker.2@student.rug.nl Supervisor: Dr. K. Scholten

(2)

i Supply Chain Resilience, learning from disruptions: A multiple case study in the water supply industry.

PREFACE

In February 2015 I started working on my master thesis of the MSc. Supply Chain Management. This journey, full of struggles has required an enormous amount of effort and time in order to get familiar with the concept of supply chain resilience (SCR) and setting up a well-structured and interesting research. Besides my own effort and time, my supervisor Dr. K. Scholten and my co-assessor Prof. dr. D. P. van Donk have put their effort and time into this research. I would especially like to thank Dr. K. Scholten for her great support during this research. Her support enabled me to get through the phases of writing a thesis in moments of insecurity and doubt. Furthermore, she also provided intellectual support during the entire process of writing this thesis. My gratitude goes to my co-assessor Prof. dr. D. P. van Donk, for his valuable support in the proposal phase of this research and for providing critical feedback on this master thesis. Company X, allowed me to investigate their supply chain (SC) and provided a place at their office for me to work. They were always willing to help and provide me the required data during the complete process. Therefore, I would like to thank all employees of Company X who were involved in this research and I would especially like to thank my three company supervisors.

(3)

ii Supply Chain Resilience, learning from disruptions: A multiple case study in the water supply industry.

ABSTRACT

While earlier research recognizes the importance of learning for supply chain resilience (SCR), literature to date has not provided additional insights on how learning actually affects SCR. By usage of a multiple case study, which makes use of quantitative and qualitative data in the context of a public utility company, this research showed empirically how learning could influence SCR to support development of SCR literature. Concluding, how learning from disruptions could improve SCR, is by the learning culture, monitoring of disruptions, the evaluation process of disruptions, implementation of improvements, implementation monitoring, and by the aims of learning. Furthermore, this research contributes to literature about SCR because it has shown that learning on supply chain (SC) level has more influence on SCR than learning on organizational level. Findings of this study provides understanding of the influence of learning on SCR and opportunities that can be used by organizations and managers in order to improve their level of SCR.

(4)

iii Supply Chain Resilience, learning from disruptions: A multiple case study in the water supply industry.

CONTENT

1 INTRODUCTION ... 1

2 THEORETICAL BACKGROUND ... 2

2.1 Supply Chain Resilience (SCR) ... 2

2.2 Learning... 4 2.3 SCR & Learning ... 6 3 METHODOLOGY ... 8 3.1 Research method ... 8 3.2 Case setting... 9 3.3 Case selection ... 10 3.4 Data collection ... 13 3.5 Data analysis... 15 4. FINDINGS ... 17 4.1 Introduction ... 17 4.2 Learning culture ... 18 4.3 Learning procedure ... 20 4.5 Aim of learning ... 25 4.6 Learning level ... 28 5. DISCUSSION ... 30 5.1 Introduction ... 30 5.2 Learning culture ... 30 5.3 Learning procedure ... 31 5.4 Aim of learning ... 33 5.5 Learning level ... 33

6. CONCLUSION & LIMITATIONS ... 34

6.1 Conclusion ... 34

6.2 Limitations & further research ... 35

REFERENCES ... 37

APPENDIX I. GENERAL QUANTITATIVE DATA ANALYSIS ... 41

APPENDIX II. GENERAL EVALUATION REPORT ANALYSIS... 50

APPENDIX III. CASE STUDY PROTOCOL ... 52

(5)

1 Supply Chain Resilience, learning from disruptions: A case study in the water supply industry

1 INTRODUCTION

The ability of a supply chain (SC) to deal with disruptions is called supply chain resilience (SCR) (Blackhurst, Dunn & Craighead, 2011; Pettit, Croxton & Fiksel, 2013; Peck, 2005) and has become an important issue for companies in order to prevent sever or long term consequences that can have financial effects or influence market positions (Ambulkar, Blackhurst & Grawe, 2015; Narasimhan & Talluri, 2009; Blackhurst et al., 2011). The level of resilience of a SC is defined as the ability to prepare for, respond to and recover from disruptions (Spiegler, Naim & Wikner, 2012) and is determined by three components; the amount of change a SC can handle, the ability to handle disruptions and the ability to learn from disruptions (Ponomarov & Holcomb, 2009). While we do know about the influence of learning on SCR from earlier research (Paton, Smith & Violanti, 2000; Hollnagel, Woods & Leveson, 2007; Ponomarov & Holcomb, 2009; Bhamra, Dani & Burnard, 2011; Scholten, Scott & Fynes, 2012), literature to date has not provided additional insights on how learning actually influences SCR. Therefore, it is not clear how SCR could be improved by learning and potential major opportunities to improve being prepared for, handling of and recovery from disruptions as a SC are left unused.

(6)

study: How can learning from disruptions improve supply chain resilience?

By making use of a multiple case study that allows to draw on qualitative and quantitative data, this research makes three core contributions. This research will contribute to current literature of SCR by providing insights how learning actually influences SCR. Secondly, this research will contribute to current literature about the influence of learning on SCR by providing insights into the effect of the level of learning on SCR improvement. The third and practical contribution of this research is that it provides insights into the influence of learning on resilience of the SC investigated, provides opportunities for improvement of their SCR and provides insights that can be used by other organizations in order to improve their SCR.

The remainder of this research is structured as follows. First, the theoretical background for the topics SCR and learning is provided. Hereafter the methodology and the results of the case study will be discussed. Thereupon, the results of this study will be summarized and propositions will be provided. Finally the conclusion and limitations of this research are provided.

2 THEORETICAL BACKGROUND

2.1 Supply Chain Resilience (SCR)

A disruption in a supply chain (SC) is an event occurring at participants of the SC, that negatively influences the SC down- and/or upstream and not just the single participant where the disruption occurred (Ambulkar et al., 2015; Craighead, Blackhurst, Rungtusanatham & Handfield, 2007). Literature provides different definitions of SCR, an overview of (SC) resilience definitions can be found in the article of Pettit, Fiksel & Croxton (2010). The definition of SCR that is used within the context of this research is; the ability of a SC to recover and return to the state of the process before, or to an even better state after occurrence of a disruption (Christopher & Peck, 2004; Jüttner & Maklan, 2011) in an allowable period of time (Brandon-Jones, Squire, Autry & Peterson, 2014). Literature to date has provided several factors that influence SCR. Ponomarov & Holcomb (2009) provide three properties of SCR based on research of Carpenter, Walker, Anderies & Abel (2001). These are:

 “The amount of change that a system can undergo while retaining the same controls on structure and function”(p.126)

(7)

 “The degree to which a system develops the capacity to learn and adopt in response to disturbances”(p.126)

Equivalent to the first two properties of Ponomarov & Holcomb (2009), Sheffi & Rice (2005) mention two ways of becoming more resilient, “creating redundancy” or by “increasing flexibility”. In a SC context, increasing flexibility improves the handling of disruptions while creating redundancy decreases the amount and chance of disruptions occurring or influencing the SC (Sheffi & Rice, 2005). Similar to this flexibility and redundancy, Wieland & Wallenburg (2013) highlight that agility and robustness influence SCR. These influences overlap with redundancy and flexibility because agility is the ability to respond and react on disruptions, while robustness is the amount of change a SC can undergo without being really disturbed. Scholten & Schilder (2015) provide an overview of factors of SCR based on literature to date concerning SCR. The factors indicated seem to have overlap (Scholten & Schilder, 2015) with each other and therefore a smaller amount of factors can be used. Scholten & Schilder (2015) argue that most of the factors can be covered by the factors indicated by Jüttner & Maklan (2011) which are; flexibility, velocity, visibility and collaboration. Because the factors of Jüttner & Maklan (2011) are widely accepted and used in literature to date, these factors will be used during this research.

Flexibility is the ability of a SC to handle, solve and exploit disruptions (Jüttner &

(8)

article of Scholten & Schilder (2015).

While the factors of Jüttner & Maklan (2011) are well recognized to enable SCR, this conceptualization neglects important conditions such as risk management, culture, learning and knowledge management. Ponomarov & Holcomb (2009) indicate with their third property of SCR that “the capacity to learn from an unexpected event and develop better preparedness fur future disruptions is a principal property of SCR” (p.137). Other literature (Paton et al., 2000; Hollnagel et al., 2007; Ponomarov & Holcomb, 2009; Bhamra et al., 2011; Scholten et al., 2012) about the influence of learning on SCR, also highlighted that learning from disruptions is required in order to continuously improve SCR but does not indicate how. In order to fully understand how learning could influence resilience, it first has to be clear what learning is. Therefore, the following section will discuss what learning actually is in more detail.

2.2 Learning

Organizations are learning when they change their future behavior in response to outcomes of their current and past performance (Bingham & Davis, 2012). Weick & Quinn (1999) mention that change in organizations can result in changes in how people do their work, organizational routines, usage of resources and structures. While these factors are found to be indicators of change in behavior, these factors on their own do not specify if the change is a result of learning. Fiol & Lyles (1985) argue that in order to say that learning occurs, two requirement have to be met; there has to be a change in systems or behavior and there has to be an association between changing past actions and the effect on future outcomes. Without this association the change is just an adaptation of new working methods (Fiol & Lyles, 1985). Ron Lipshitz & Popper (2006) indicate three steps of learning from disruptions; “what happened” (p.1077), “what went wrong” (p.1077) and “how can we do better next time”(p.1077). These steps indicate monitoring, analysis of the disruption and usage of this disruption in order to improve, so there has to be a causal relation between change of past actions and results in the future. The requirements of Fiol & Lyles (1985) are consistent with the three steps of learning from Ron et al. (2006), but Ron et al. (2006) pay more attention to the collecting of information to learn from.

(9)

to achieve what. Furthermore, because monitoring was found to provide opportunities to learn from, monitoring is used during this research as an influences on learning. Learning as a system can take place on different levels. An overview of literature focusing on these different levels is shown in table 2.1 below.

Literature to date as shown in table 2.1, argues that individuals contribute to organizational learning (Fiol & Lyles, 1985; Dodgson, 1993; Jain & Moreno, 2015). However, organizational learning was found to be depended on the learning orientation of the firm and this orientation can be explained as the culture, effort put into and attitude towards learning inside the organization (Yeung, Lai & Yee, 2007; Weick & Quinn, 1999). As shown in table 2.1, when looking at a higher level, learning as an organization is influenced by the learning of the SC the organization is part of (Dodgson, 1993; Peterson, 2002; Spekman et al., 2002; Bessant, Kaplinsky & Lamming, 2003; Jain & Moreno, 2015; Mentzer, DeWitt, Keebler, Min, Nix, Smith, & Zacharia, 2001). This suggests that SC members influence each other’s learning processes. Peterson (2002) states that SC learning is the combination of organizational learning and an integrated SC. Furthermore, he points out that culture as mentioned before for single organizations, is of the same importance for SC learning and influences the willingness to collaborate of participants of the SC. Learning on a SC level combines the benefits of the integration of SC participants with knowledge management practices up- and downstream the SC and therefore provides more benefits than a single learning organization or a “normal” SC (Peterson, 2002). Besides monitoring, the learning culture is used as one of the influences on learning during this research because it was found to have an important influence on the willingness to learn and influences improvement on SC level.

(10)

TABLE 2.1 - Learning levels

2.3 SCR & Learning

In table 2.2 an overview of literature concerning the influence of learning on (SC) resilience is provided in order to create insights in what is known about this relation to date. Literature as shown in table 2.2 highlights that learning from disruptions is required in order to continuous improve SCR (Paton et al., 2000; Hollnagel et al., 2007; Ponomarov & Holcomb, 2009; Bhamra et al., 2011; Scholten et al., 2012). This is particularly important in relation to risk and disruptions because learning could reduce the chance of risks and disruptions occurring in the future (Pidgeon, 1991; McDaniels & Gregory, 2004). Therefore, this can be linked to the SCR factor flexibility (which is about absorbing disruptions without letting them disrupt the SC) of Jüttner & Maklan (2011), while the performance improvement as a result of learning can be linked to the SCR factor velocity (which is about the speed of solving disruptions) of Jüttner & Maklan (2011). Hollnagel et al. (2007) indicate that besides the importance of learning from disruptions in order to become highly resilient, also the continuous learning from so called “near miss incidents” (p.329) and disruptions should be seen as opportunities for improvement.

Author(s) Levels

Fiol & Lyles (1985)

• Individual level • Organizational level

Learning as an organization is not just the sum of all inidividual learning, but individual learning affects and promotes learning as an organization.

Dodgson (1993) • Learning as an organization is reliant on its members (individuals) • The learning culture should be extended to the entire supply chain Peterson (2002) • Learning as an organization is influenced by learning of the supply

chain • Learning on the supply chain level is explained as, learning of all members of the chain.

Spekman, Spear, & Kamauff, 2002

• Learning as a supply chain can improve effectiveness of the individual members of the chain.

Bessant, Kaplinsky & Lamming (2003)

• Learning as an organization does not stop at a single firm because firms are dependent of participants in the whole chain. • Learning on the level of supply chain can influence the performance of the single firms of the chain.

Jain & Moreno (2015) Five levels of learning as organizations • Learning as an individual.

• Learning as a group/team. • Learning as an organization • Learning from other organizations

• Recognizing the effect of one level on another

(11)

Pidgeon (1991) argues that monitoring makes it possible to see these near misses. Without monitoring the near misses were never detected and the opportunity to learn from was lost (Pidgeon, 1991). Therefore, the SC should be monitored in order to indicate weak points or possible disruptions (Blackhurst et al., 2011) in order to respond to (Ponomarov & Holcomb, 2009). These near misses can indicate that SC are resilient and this can be linked to the SCR factor flexibility of Jüttner & Maklan (2011) again because they did not disturbed the SC but were absorbed. The monitoring which was found to be one of the determinants of learning (Ron et al., 2006) makes disruptions visible and can be linked to the SCR factor visibility of Jüttner & Maklan (2011).

Furthermore, literature concerning SC learning highlights that the internal culture influences the willingness to learn (Weick & Quinn, 1999;Yeung, Lai & Yee, 2007) and the willingness to collaborate as a SC (Peterson, 2002). This culture, attitude, willingness to learn and the degree information is shared (knowledge management) could be linked to the SCR factors collaboration and visibility of Jüttner & Maklan (2011).

TABLE 2.2 - Influence of learning on resilience

Because this research is SC oriented, it also focuses on resilience on SC level, but because SC’s consist of multiple organizations that influence the learning of the SC, this

Author(s) Influence of learning on resilience Type of research

Paton, Smith & Violanti (2000)

• Resilience is a proactive process of learning, growth and self

improvement with the aim of improved functioning in the future. • Capacity of individuals and experiences of the past are crucial for

this.

Literature review

Hollnagel, Woods & Leveson (2007)

• Flexibility and the adaptive capacity are influences on resilience which are required to react on changing environments. • Monitoring, learning and improving are found to be crucial for resilience. • Besides the importance of learning from disruptions also the continuous learning from so called “near miss incidents” (p.329) is found to be important. Literature review in combination with emperical findings of practioners and safety managers

Ponomarov & Holcomb (2009)

• The capability of systems to learn from disruptions in order to improve handling disruptions in the future and improve preparedness for disruptions is one of the aspects of resilience.

Literature review, 400 articles Bhamra, Dani &

Burnard (2011)

• The learning and adaptive capacity improve the ability of companies to react on a changing environment

Literature review, focussing on SME´s Scholten, Scott & Fynes

(2012)

• Learning from disruptions is crucial for organizations in order to improve dealing with disruptions and recover from disruptions. • Learning from disruptions and the organizations experience with

learning from disruptions can provide advantages in the future. • Disruptions should be seen as opportunities to learn positive

lessons out of.

Literature and emperical findings collected with a case study

(12)

research will focus on both organizational and SC learning. While the factors of SCR and SC learning seem to be closely related to each other and monitoring, learning and improving as a SC seems to influence SCR, the link between them is still not clarified. This research will make use of the conceptual model shown in figure 2.1 in order to provide more in-depth insights of the underlying relationships.

FIGURE 2.1 - Conceptual model

3 METHODOLOGY

3.1 Research method

(13)

practice of it(Voss, 2009; Yin, 2009), which is important in order to get representative view of the actual situation. This is required for this research in order to see the underlying influences of learning on SCR in practice.The main research question, How can learning from disruptions improve supply chain resilience, is a “how” question. Case studies are particularly suitable for answering “how” questions (Voss, 2009; Yin, 2009). Case studies allow making use of the combination of multiple research methods, for this research there will be made use both quantitative and qualitative data (Voss, 2009; Yin, 2009; Eisenhardt, 1989). All these factors make usage of case study method in this research very appropriate and therefore, a case study is chosen as the method that will be applied. The unit of analysis of this research is disruptions, this will make it possible to investigate the process of leaning the SC uses to learn from disruptions in order to improve its SCR in multiple cases. Multiple cases of disruptions provide a more complete and robust view of the topic investigated compared to a single case and amplify external validity (Voss, 2009; Yin, 2009; Eisenhardt & Graebner, 2007).

3.2 Case setting

This research consists of a multiple case study conducted at Company X who both produces and distributes its products. In the Netherlands there are ten public water suppliers which are regulated by the Dutch government (Rijksoverheid, 2014). Private water suppliers are not allowed by the government. Public sector companies are found to be less innovative, struggle more with learning and improving processes compared to private companies (Borins, 2001; Bate & Robert, 2002). Therefore, Company X is particularly appropriate for investigating learning inside a SC because public sector companies seem to have more opportunities for learning compared to private companies.These factors make Company X a representative case for investigating the influence of learning on SCR (Yin, 2009).

Water is found to be one of the basic necessities for hygiene, health and growth of a society and availability is therefore crucial (World Health Organization, 2006). Therefore, the government puts high pressure on Company X to ensure availability of water to its customers (Rijksoverheid, 2014). Because of the high consequences of not being able to deliver, the change of and effects of disruptions should be minimized as much as possible in the Company X case and SCR is of high importance.

(14)

can be seen as two different organizations inside Company X. Because Company X also both produces and distributes its products, disruptions at this company can occurs in both the production and distribution of water, of two different sub-organizations (municipal and industrial). Therefore, this case company allows to investigate resilience in a SC context and to see the dyadic relations. In order to see if disruptions are handled/used differently in order to learn from in the SC of Company X and to see if there was learned on organizational or on SC level, both MWS and IWS were part of the scope.

3.3 Case selection

Available data from databases of Company X concerning disruptions that occurred in the SC of Company X was analyzed in order to create insights in the amount of, nature and extent of disruptions. This data from the databases contained the following information, when did a disruption occur, duration of the disruption, location of the disruption, detailed cause of the disruption information and who handled the disruption. Two overall causes of disruptions could be identified, pipeline breakdowns and pressure problems (pumps). More detailed information of the database analysis can be found in Appendix I.

Besides this data, there were disruption evaluation reports available that showed how Company X currently uses disruptions in order to continuously improve. The company investigated evaluated disruptions if they met predetermined criteria (see Appendix II). The evaluations of disruptions provide suggestions for improvement (SFI’s) and after this evaluation process, these SFI’s are implemented by the company investigated with the aim of continuous improvement. Because the company investigated already was extensively evaluating disruptions in order to continuously improve dealing with disruptions, the disruptions that were evaluated were ideal cases in order to see how occurred disruptions were used with the aim of continuous improvement and if/how dealing with disruptions was improved. Evaluation reports were available for both MWS and IWS and for both pipeline breakdowns and pressure problems.

Within the evaluation reports a distinction could be made between two different kind of disruptions; incidents and crisis situations. Incidents are situations that do not require a crisis team and crisis situations are incidents that require a crisis team because of the criticalness of the situation. In the last four years two crisis situation occurred and both were investigated in order to indicate the difference between incidents and crisis situations.

(15)

criteria for learning, association of effect of change), on the fact if they were disruptions of the different sub-organizations (in order to see if learning occurred at organizational or at SC level) and on the criticalness of the disruptions to see if this has an influence. Two cases for IWS, MWS incidents and MWS crisis situations were selected (six in total). Investigating six cases is found to be sufficient in order to obtain convincing findings (Eisenhardt, 1989). For all three groups of cases, the most recent cases were selected that were evaluated in order to increase the chance that data concerning these cases was available and possible interviewees could remember the disruptions of the cases.

(16)

(17)

3.4 Data collection

Available database data concerning disruptions that occurred from 1 January 2010 till 31 December 2014 was analyzed. Besides this database data, evaluation reports and action lists of disruptions that occurred from January 2010 and 31 December 2014 were investigated. This quantitative data was used to create insights into the factors flexibility, visibility, velocity, collaboration, learning, level of leaning, monitoring, causes of disruptions and dealing with disruptions. The entire procedure of learning from disruptions was investigated and an overview of this procedure is shown in appendix II. System monitoring and this learning procedure (evaluation process, implementation and implementation monitoring) were investigated for all six cases (disruptions) to see how and if learning occurred. The departments that were involved in the selected cases for in-depth research were identified in order to create an overview of potential interviewees. If information from the quantitative data was unclear or incomplete, semi-structured interviews with (production) managers and employees responsible for handling disruptions and maintenance were conducted. The interviewed managers were chosen because they possessed the required knowledge for answering the questions and because they had access to necessary resources, like experience with dealing with disruptions. The interviewed employees responsible for handling disruptions and maintenance were chosen because they directly had to deal with the occurred disruptions and provide different insights compared to the insights obtained from the interviews with (production) managers who were more appropriate for clarification of more general ambiguities. An overview of the interviews, durations and locations of the interviews is shown in table 3.2 below. The interviews were also used to create more insights in monitoring, culture, documentation and aftercare of the occurred disruption after the disruption was solved.

(18)

TABLE 3.2 – Interviews

An overview on data collection sources can be found in table 3.1. For more information about the interviews, see the case protocol in Appendix III.

(19)

FIGURE 3.1 – Research method overview

3.5 Data analysis

In order to say learning has influence on SCR two thing are required. There has to occur learning and there has to be an influence of this learning on SCR. Therefore, both requirements are investigated. The first part of the analysis focused on the learning culture and on how the procedure of learning from the investigated disruptions took place and if this influenced and led to SCR. The second part of the analysis focused on the influence of the SFI’s of the cases on SCR and the influence of the learning levels. Definitions of important factors for this research can be found in appendix IV.

Analysis of the interview results and coding of the results was done in four steps. First,

Quantitative Data Analysis:

Database analysis

Disruption evaluation general analysis

Case Selection (six cases of three groups):

Two customers types (Municipal and Industrial) Two disruption types (Incident and Crisis Situation)

Three disruption causes (Pipeline breakdown, Pressure problem and Quality problem) Most recent cases per group

Quantitative Data Analysis

In depth analyis of evalution reports (of the cases)

Analysis of evaluation report results and usage (actionlists)

Analysis of changes in manuals and/or documents

Implementation of learned knowledge

Additional Interviews/Qualitative data

Collection of additional information to support findings from quantiative data

Monitoring of implementation

Interviews according to the protocol and based on the factors investigated during this study

Cross-case comparison

(20)

the data was reduced to relevant quotes/answers for this research. Secondly, they were coded into more general categories/topics (descriptive codes) in order to see how Company X tried to learn and stimulated learning. Thirdly, they were coded based on how these efforts of learning could influence learning, for example “stimulate learning process” or “stimulate learning culture”. Fourthly, they were coded on if they influenced change in behavior or systems or association of change which are the criteria for learning used during this research. The analysis of the interview results made it possible to see what influences learning, learning culture, create insights into the learning and implementation process and to see their influences on SCR. At the start of the analysis, for all cases the influences on learning were indicated and compared with each other to indicate what exactly influences the need for change. Thereupon, the influence of the learning culture on SCR was indicated. After that, for all six cases was analyzed how the association of the effect of change was created and the results were compared with each other to see the overall influence of this on learning and SCR. Hereafter, the influences on change in behavior or systems were analyzed for all six cases and compared again. These steps were taken in order to indicate what influences learning and therefore can influence the influence of learning on SCR.

(21)

influence of the learning level. This made it possible to indicate the underlying relationships between learning and SCR and to see the relationships between the level of learning and SCR. The analysis first focused on the different factors of SCR that were affected by the SFI’s to identify patterns. After this, other findings from the analysis were highlighted and this was followed by the findings about the influence of learning on SCR. Thereupon, the findings of the influence of the level of learning of the cases were shown.

The analysis and coding scheme of the interviews highlighted how learning was influenced by different factors and therefore really focused on the influences on learning. These factors that affect learning could therefore influence SCR again. The analysis and coding scheme of the SFI’s indicated how SCR was influenced by learning and the influence of learning levels. Both schemes supported each other by providing insights in order to answer the research question. For example, the coding scheme of the interviews showed what influenced change in behavior or systems and therefore SCR, and the coding scheme of the SFI’s showed how actual change in behavior or systems influenced SCR. Appendix IV provides examples of the coding schemes.

4. FINDINGS

4.1 Introduction

The analysis of all the collected data made it possible to indicate the role of monitoring, see what influences the learning culture and the influences on the learning criteria’s; association of the effect of change and change in behavior or systems. Most important, it made it possible to show how learning could influence SCR and which level of learning had the most influence on SCR. The analysis of the interview results and the learning procedure showed that monitoring, the external stakeholders, the learning culture and the learning procedure influenced how active and effective a SC learns and how active and effective SCR was improved. Furthermore, the analysis of the SFI’s showed that the aims of learnings influenced the factors SCR and that SC learning has more influence on SCR than organizational learning.

(22)

4.2 Learning culture

Analysis of interviews results showed how the learning culture affected the influence of learning on SCR. Table 4.1 shows the differences between attitudes and efforts of learning of the sub-organizations IWS and MWS. An interviewee (#1) of the IWS department indicated that “the mindset of most employees of the IWS department is focused on continuous improvement” (#1) and that therefore this part of the SC is constantly learning. This active learning culture was found to be stimulated by the customers. As an interviewee(#1) indicated, “in the IWS sector there is in most cases a direct consequence of disruptions for customers and therefore the need for optimization is also higher”(#1). Table 4.1 shows that the efforts and attitudes toward learning of the MWS department differ from the IWS department and that learning was more top-down organized compared to the IWS department, where learning was both top-down and bottom-up organized. As shown in table 4.1, both IWS and MWS made use of periodical meetings in order to stimulate learning between employees and they make use of evaluations in order learn from more severe disruptions. Furthermore, interviews highlighted that the MWS department also participates in projects in order to improve reacting on its environment, which again indicates the environmental influence. Besides that, an interviewee (#5) indicated that, “nowadays, the media is way quicker aware of disruptions that occurred than years ago and Company X wants to minimize this negative media attention. These changes in the environment force Company X to react more quickly on disruptions” (#5). This shows that the media also influences the need for change. While this shows how the sub-organizations stimulate learning, it does not show if they learn together.

(23)

implies learning on SC level. However, most of the learnings that were SC orientated, focused on a part of the SC and not on the entire SC. So even the learnings that were SC orientated were in most cases partly SC orientated. Even though there were some indicators of SC learning, in this study most learning occurred on organizational level.

It was found that while both sub-organizations had a strong learning culture, the learning cultures differed throughout the SC. A strong SC learning culture could not be identified and information sharing was limited. The difference in learning cultures was found to be mainly caused by the fact that the sub-organizations of Company X deliver different types of products and the customers influence the need for change. Besides that, increased media attention forced Company X to improve its systems and reduce the chance of and durations of disruptions which could be linked to SCR. Furthermore, the learning culture influenced how active there was learned, if the learnings were shared (which could be linked to the SCR factor collaboration) and therefore if learnings were available to all members of the SC (which could be linked to the SCR factor visibility) so other members of the chain could also use the learnings to improve (could be linked to all SCR factors). The environmental stakeholders were found to stimulate the need for SCR, the learning culture was found to affect learning and this influence could be linked to factors of SCR.

TABLE 4.1 – Learning efforts & attitudes

Learning culture

Sub-organization

IWS C1 & C2 MWS C3, C4, C5 & C6 Active learning Active learning culture employees.

Mindset employees focused on learning (#1)

Top-down and bottom-up learning

Top-down learning

Meetings for information sharing

Yes, all employees are present in order to learn from each other. (#1) Talk about disruptions during these meetings (#1)

Yes, to discuss occurred disruptions that met not met the criteria for detailed evaluation. (#2, #3)

Participate in meetings with other water suppliers to learn (#4)

Evaluating of disruptions Yes, evaluating disruptions that have great impact on customers. (#1)

Yes, evaluating disruptions that have great impact on customers (#2, #3, #4 & #5)

Participate in projects Yes, to stay up to date and keep reacting on the environment (#3 & #5)

Training sessions Yes, practice dealing with disruptions and involve employees (#4 & #5)) Participate in training sessions of other water suppliers (#4)

Unexpected quality checks

(24)

4.3 Learning procedure

In order to see the influence of the learning process on learning from disruptions to improve SCR, the learning process was analyzed. The process of learning consists of monitoring disruptions to learn from, the learning procedure (evaluating disruptions to identify SFI’s for implementation, implementation and implementation monitoring). Figure II.1 in appendix II gives an overview of this procedure of learning from disruptions. All steps were analyzed to indicate their influence on SCR.

Monitoring

Analysis of interviews showed that Company X monitors its SC and keeps track of all disruptions that occur. Besides that, there are sensors installed on certain location in order to monitor water quality and to react on deviations as quickly as possible. After the disruption of case two (C2), there were sensors installed on the pressure pumps of IWS in order to also monitor this part of the SC. In case of a pipeline breakdown, sensors also indicate where abnormal water loss occurs. This improved monitoring as a result of learning from disruptions, provides more insights in the SC and could be linked to the factor visibility of SCR. The improved visibility allows to faster solve disruptions and near misses that occurred which could be linked to the SCR factors flexibility and velocity. Monitoring systems to get insights in near misses in order to learn from, was illustrated by case four (C4), where the monitored system indicated an increase in water pressure which was linked afterwards to a series of pipeline breakdowns that occurred after the increase of pressure. After the C4 disruption, the relationship of this increase of water pressure and pipeline breakdowns was investigated in order to see if there was a connection and prevent possible disruptions in the future.

This shows that, monitoring disruptions provides opportunities to learn from and improve SCR. Furthermore, monitoring the system provides opportunities to link disruptions to changes in systems and to improve associations of change. It was also found that the monitoring improved visibility of the system and improved reacting on deviations in the SC which influenced the SCR factors flexibility and velocity.

Evaluating & Identifying SFI’s

(25)

Associations can results from meetings with other water suppliers as indicated by an interviewee (#4), but Company X also has its own procedure in order to learn from disruptions. Monitored disruptions that were not evaluated with an evaluation form, were discussed at periodical meetings “in order to see how there was dealt with the disruptions and what possibly can be improved” (#3). Disruptions that had more impact on customers, were extensively “evaluated and used for continuous improvement” (#4). Company X makes use of a standard evaluation form in order to make sure that all evaluations are complete, to reduce time and effort required to evaluate and that it is always clear where to find what in an evaluation report. Table 4.2 shows that not all cases were evaluated with this standard evaluation method. C2 was evaluated with a deviating evaluation form. An interviewee (#1) mentioned that the evaluation form of C2 was used by the IWS department because they did not knew that the standard form was available. Table 4.2 shows that also for both case five (C5) & case six (C6), deviating evaluation forms were used. An interviewee (#4) indicated that normally evaluations were focusing on all employees/partners involved in a disruption but for C5, the “evaluation report was actually not evaluated with the standard form. For this evaluation, all crisis team members were asked how the crisis team functioned during the meetings and if they noticed extraordinary things” (#4). This evaluation was focusing on the crisis team and not on all participants. This interviewee (#4) also highlighted that the “normal” evaluation should had been used because the evaluation report of C5 was an unorganized list of findings, did not deliver clear SFI’s and afterwards SFI’s still had to be identified which resulted in extra work. C6 was evaluated with a deviating form because in 2010 there was no standard form. The evaluation was unorganized and SFI’s were spread all over the report. The evaluation reports provided SFI’s that were transferred (if agreed by the department that takes care of disruptions) to action lists for the actual implementation. Table 4.2 shows the amount of SFI’s from the evaluation reports of the cases investigated and the amount of SFI’s on the action lists. Table 4.2 however, indicates some deviations.

(26)

reasons why evaluations were discussed at periodical meetings.C6 was the oldest disruption investigated. In 2010 there were no action lists where the SFI’s from the evaluation report were transferred to and therefore it was not clear what happened with the SFI’s after setting up the evaluation report. C6 clearly indicates that the procedure of evaluating has improved throughout the years. Furthermore, table 4.2 shows that the criticality of the disruption was found to have influence on the evaluation process and amount of learnings. C1 & C2 were IWS disruption that directly influenced customers. C3 & C4 were MWS disruptions that were less severe. C5 & C6 were MWS crisis situations where high amounts of customers were affected. As shown in table 4.2 the higher the criticality of the disruption, the more SFI’s were identified from the evaluations and the more was tried to improve SCR. It also shows that the more critical a disruption was, the higher the need for SCR.

Analysis of the evaluation methods and action list usage of the cases showed that learning from disruptions was depended on the learning process. Some learnings from evaluations were not transferred to the implementation phase because of unstructured and non-uniform evaluation methods. Also the lack of action lists for implementation resulted in the fact that it was not clear what was done with the learnings from evaluations. Besides these findings, the criticality of the disruptions influenced the need for SCR and how active there was learned from the disruptions. Furthermore, it was found that the phase between evaluating (identifying SFI’s) and transferring SFI’s to the implementation phase was crucial for learning to influence SCR. If disruptions were evaluated the association of the effect of change was created, but if the identified SFI’s were not implemented, no change in behavior or systems occurred and learning did not influence SCR.

TABLE 4.2 – Evaluation method & implementation findings

Cases Evaluation method #SFI’s # SFI’s on action lists # SFI’s Implemented

C1 Normal 7 4 4

C2 Deviating, form of industrial water

department

6 6 6

C3 Normal 3 2 2

C4 Normal 2 3 3

C5 Deviating, different evaluation form 20 20 17

C6 Deviating, old evaluation form 22 There were no action

lists in 2010

(27)

Implementation and implementation monitoring.

(28)

implementation was still in progress, the second SFI was not implemented successfully enough and the third one should had been implemented but it was not clear why this was not done.

TABLE 4.3 – Implementation

Besides the chance of human mistakes when implementing SFI’s, the interviews also provided insights into other pitfalls of implementation. If the action lists for monitoring progress of implementation were not available or up to date, status of implementation was not clear. Furthermore, an interviewee (#5) of C6 indicated that, “small manuals/working instructions can be used but large documents are not used by employees in reality” (#5). A large amount of SFI’s from the evaluation reports were adjustments in manuals/working instructions and the effectiveness of this can be limited. An interviewee (#4) mentioned that, “one of the things that Company X tries to do when setting up SFI’s is formulating them SMART”(#4) which should counteract, “SFI’s which are not that tangible, for example adjusting working methods, are way more difficult to implement”(#4). This interviewee (#4) also indicated that, “optimization of intangible things is a continuous improvement process which requires practice and continuous evaluation of performance/functioning” (#4). Intangibility is found to have influence on the implementation process and should be avoided and if not possible, be accepted. During this study also database data of monitored disruptions was analyzed to compare durations of disruptions throughout the years in order to indicate improved SCR (velocity). However, an interviewee (#2) of C3 mentioned that, “every disruption can be different. This can result in different and new SFI’s, therefore SFI’s that are implemented are no guarantee for faster disruption solving in the future” (#2). This implies that even when SCR was improved this was not always measurable with durations.

Analysis of the implementation process and implementation monitoring showed that these phases were crucial in order to let SFI’s result in change in behavior or systems and let learning improve SCR. If learnings from evaluations were not implemented and used, there was no change in behavior or systems, no learning and no influence of learning on SCR.

Implementation process

Sub-organization

IWS C1 & C2 MWS C3, C4, C5 & C6 Implementation method Active implementation Adjustments in documents

Explanation change to relevant employees

Implementation monitoring method

Direct monitoring Training sessions

(29)

4.5 Aim of learning

The analysis of the SFI’s of the cases showed that almost all SFI’s were related to multiple factors of SCR. SFI’s related to collaboration were, except for a few SFI’s, also related to other factors of SCR. Furthermore, SFI’s related to the factor velocity, were except for one SFI of the cases, also related to the factor visibility. This was caused by the fact that making information visible could improve the speed of handling disruptions (Velocity). For example, “when the transport pipeline breaks down, mechanics should consult the manual for this type of disruption and this manual should clearly indicate which material is required”(SFI 17, C6). By creating a manual that indicates how to solve a specific disruption and what is required for this, mechanics could work more effectively and efficiently which could improve the speed of handling the disruption. Also when different mechanics are used in shifts the knowhow is available because the manual is visible to all employees. Another example, “provide extra information on time concerning the length and pressure requirements in the right computer program” (SFI 1, C1), which is about information being available (Visibility), has to be shared on time between different SC members (Collaboration) and it improves the speed of handling disruptions (Velocity). Categorization of the SFI’s based on the aims behind the SFI’s highlighted the same interrelationships between the SCR factors and showed that the aims of learnings were found to be related to the factors of SCR.

The factor flexibility was mainly influenced by learnings with the aim of reducing the risk of disruptions occurring by preparing for disruptions. Velocity was mainly influenced by learnings with the aim of improving the speed of solving disruptions and making clear how to solve disruptions. The factor visibility was found to be closely related to velocity and the learning aims also indicated this because, visibility was mainly influenced by learnings with the aim of making information available in order to improve the speed of solving disruptions. Collaboration was mainly influenced by learnings who also influenced the other three factors of SCR, but collaboration adds the involvement of suppliers, closely related companies and other chain members. Furthermore, table 4.4 shows that the aims behind single learnings could influence multiple factors of resilience and also shows interrelationships between the SCR factors.

(30)

factors of SCR. This indicated that learning can influence SCR by the aim of learning and therefore, this shows that by learning from past disruptions and near misses in order to identify and implement SFI’s, factors of SCR can be influenced and preparedness for, handling of and recovery from disruptions can be improved. In order to make more clear how learning from disruptions could improve SCR, two examples from the cases will now be provided.

(31)

(32)

4.6 Learning level

(33)

affected by SC learning. Besides that, table 4.5 showed that C1, C2 and C3 were more focused on organizational learning and that C4, C5 and C6 were also more focusing on SC learning. Table 4.6 shows that all cases score relatively the same for organizational learning but on the level of SC learning, C4, C5 and C6 had more influence on the factors flexibility and collaboration compared to C1, C2 and C3. This again shows that learning on SC level influences more factors of SCR than learnings on organizational level which also supports that SC learning has more influence on SCR than organizational learning.

TABLE 4.5 – Learning levels & SCR

TABLE 4.6 – Learning levels specific & SCR Learning

level

Effect on learning level

IWS MWS

C1 C2 C3 C4 C5 C6

Organizational learning

High High Medium/High Medium Medium/High Medium/High

Supply chain learning.

Low Low Low/Medium Medium Medium Medium

Factors of SCR

Effect on SCR

C1 C2 C3 C4 C5 C6

Flexibility Low Medium/High Low/Medium Medium Low/Medium Low/Medium Velocity High Low/Medium High High High High

Visibility High Medium/High High High High High

Collaboration Medium Low Low Medium Medium Low/Medium

Factors of SCR

Effect of learning level of SFI on SCR factors Organizational learning

IWS MWS

C1 C2 C3 C4 C5 C6

Flexibility Low Medium/High Medium Low Low Low Velocity Medium/High Low Medium/High Medium High High Visibility High High High Medium High High Collaboration Low/Medium Low Low Low Low Low

Supply chain learning

Factors of SCR C1 C2 C3 C4 C5 C6

Flexibility Low Medium Low Medium Medium Medium Velocity Medium Medium High Medium High Medium/High Visibility Medium Low High Medium High High

(34)

5. DISCUSSION

5.1 Introduction

The findings of this study showed how adjusting behavior and/or systems based on learnings from evaluated disruptions from the past, as indicated by literature (Fiol & Lyles, 1985; Ron et al., 2006) can make it possible to improve SCR as mentioned by literature (Paton et al., 2000; Hollnagel et al., 2007; Ponomarov & Holcomb, 2009; Bhamra et al., 2011; Scholten et al., 2012) in order to be more prepared for, improve solving of and recovery from disruptions in the future (Christopher & Peck, 2004; Jüttner & Maklan, 2011). The analysis of the findings made it possible to answer the question, how learning could improve SCR (Ponomarov & Holcomb, 2009) and allowed to indicate which level of learning influences SCR the most (Bhamra et al., 2011). The findings of this study showed that the learning culture, monitoring disruptions, the evaluation process, implementation of learnings, implementation monitoring and the aims of learning, influenced how learning could improve SCR. Furthermore, SC learning was found to have more influence on SCR than organizational learning. This section will focus on the interpretation of the findings of this research in more detail and will provide empirical propositions based on these findings.

5.2 Learning culture

(35)

this research showed that the learning culture influenced how learning could improve SCR and contributes to SCR literature (Ponomarov & Holcomb, 2009). Based on these findings, the following propositions were identified.

Proposition 1: The learning culture influences SCR improvement

P1a: The learning culture influences the SCR factors visibility and collaboration which influences the SCR factors velocity and flexibility and therefore the learning culture influences SCR improvement.

P1b: The environmental stakeholders determine the need for SCR.

5.3 Learning procedure

This study has shown that the different phases of the learning process influenced how learning from disruptions could improve SCR. As mentioned by Blackhurst et al. (2011), monitoring is important in order to indicate weak points or possible disruptions in the SC. This study showed, in line with Hollnagel et al. (2007), that for improvement of processes the state of the processes should be monitored. It was found that monitoring makes it possible to create the association between changing past actions and the effect on future outcomes (Fiol & Lyles, 1985; Ron et al., 2006) and provides opportunities to link disruptions to changes in systems. Furthermore, it was found that without monitoring the near misses were never detected and the opportunity to learn from was lost (Pidgeon, 1991). Besides that, it was also found that monitoring improved visibility of the systems (related to the SCR factor visibility) and improved reacting on deviations in the SC, which influenced the factors flexibility and velocity of SCR. This shows that, monitoring disruptions provides opportunities to learn from and improve SCR. This contributes to the gap in literature indicated by Ponomarov & Holcomb (2009) about how learning could actually improve SCR. Based on these findings, the following propositions were identified.

Proposition 2: Monitoring disruptions makes it possible to improve SCR because it influences the SCR factors flexibility, velocity and visibility and provides opportunities for improvement.

P2a: Monitoring makes it possible to link disruptions with changes in the systems, learn how to prevent causers of disruptions and therefore improve SCR.

(36)

better react on disruptions in the future, which is in line with Ambulkar et al. (2015). The second criteria for learning of Fiol & Lyles (1985) and Ron et al. (2006), indicated that change in behavior or systems was required for learning which highlights that only evaluating and learning how to improve processes is not enough to actually improve SCR. It was found that if disruptions were evaluated, the association of the effect of change was created but if the identified SFI’s were not implemented sufficiently, no change in behavior or systems occurred and learning did not influence SCR. Therefore, in order to make sure that change in behavior or systems, with aim of improvement occurs, the implementation process and implementation monitoring are important. This contributes to the gap in literature indicated by Ponomarov & Holcomb (2009) about how learning could actually improve SCR.

It was also found that the evaluation processes of the cases investigated, which were used in order to learn, differed and that this influenced the results of the learning procedure. The evaluation methods and action list usage of the six cases, indicated that learnings from evaluations were depended on the learning procedure and to effectively learn as a SC (Spekman et al., 2002) a uniform learning procedure was required.

Besides that, the analysis of the findings of this research showed that the criticality of disruption influences the need for SCR. The more a disruption was influencing customers, the more extensive the learnings procedure was performed and the more was learned. This again indicates the influence of the environmental stakeholders on the learning process. This study has shown that differences in learning methods of chain members influences SC learning which is in line with the findings of Peterson (2002) and Dodgson (1993). Because chain members used different learning methods they focused more on organizational level than on SC level, which influenced improvement of SCR and contributes to the gap in literature indicated by Bhamra et al. (2011) about the influence of the learning level on SCR.

Proposition 3: The learning procedure (evaluating, implementation and implementation monitoring) influences how active and effectively SCR is improved.

P3a: Improvement of SCR by learning, only occurs if implementation and monitoring of implementation take place.

(37)

P3c: Presence of a uniform learning method for all members of the SC and sharing of learnings, influences if there is learned on SC level or on organizational level and influences if SCR improves on SC level or on organizational level.

5.4 Aim of learning

This study showed that SFI’s were in almost all cases related to multiple factors of SCR. SFI’s related to the factor velocity were in almost all cases also related to the factor visibility which is in line with the results of Wieland & Wallenburg (2013) and supports their research with empirical findings. SFI’s that were related to collaboration were almost always also related to other factors of SCR, which is in line with the results of Scholten & Schilder (2015). The analysis of the findings of this study showed that the aims of the learnings were to improve specific aspects and these could be linked to the factors of SCR. The aims of the learnings were also found to influence multiple factors of resilience in most of the cases. It was shown that the aims of learnings influenced the factors of SCR that were improved by the implementation of the learnings and therefore SCR was also improved. This shows that by learning from past disruptions and near misses in order to identify and implement SFI’s, factors of SCR can be influenced and therefore preparedness for, handling of and recovering from disruptions can be improved. This contributes to the gap in literature indicated by Ponomarov & Holcomb (2009) about how learning could actually improve SCR. Based on the findings, the following propositions were identified.

Proposition 4: Factors of SCR are influenced by the aim of why a SC learns.

P4a: Aims on learnings can influence multiple factors of resilience

5.5 Learning level

(38)

this research pointed out that SC learning influences more factors of SCR compared to organizational learning, therefore indicates that SC learning has more influence on SCR and contributes to literature about the influence of learning on SCR (Bhamra et al., 2011). Based on these findings, the following propositions were identified.

Proposition 5: SC learning influences more factors of SCR than organizational learning and therefore, SC learning has more influence on SCR than organizational learning.

P5a: The SCR factors collaboration and flexibility are more influenced by SC learning than organizational learning

P5b: Organizational learning focusses mainly on the SCR factors velocity and visibility

6. CONCLUSION & LIMITATIONS

6.1 Conclusion

This multiple case study provides new interesting insights and contributes to previous studies about the influence of learning on SCR. While learning has been indicated to influence SCR, literature to date did not indicate how learning from disruptions could actually improve SCR. Furthermore, the effect of the level of leaning on SCR improvement and which level of learning influenced SCR the most, was still found to be unknown. This research has shown how learning could improve SCR and which learning level influences SCR the most. Furthermore it has provided other interesting findings to support development of SCR literature.

(39)

An additional finding of this research was that the more customers were influenced by a disruption and the more critical a disruption was, the more was learned and the more SCR was improved. This showed that environmental stakeholders influence the need for SCR and criticality of disruptions influences the need for learning and SCR improvement. A second additional finding is that a uniform learning method inside the supply chain could improve learning on SC level and learning on SC level leads to more SCR improvement than organizational learning and therefore influences SCR. Another additional finding is that monitoring makes it possible to link disruptions with changes in the systems at the time the disruptions occurred. This makes it possible to learn how to prevent causers of disruptions and therefore improves SCR.

Besides these theoretical contributions, the practical contribution of this research is that it provided insights into the influence of learning on SCR of the company investigated and provides opportunities for improvement of their level of SCR. These insights and the propositions can be used by other organizations in order to improve their level of SCR. 6.2 Limitations & further research

(40)

(41)

REFERENCES

Articles:

Ambulkar, S., Blackhurst, J., & Grawe, S. (2015). Firm's resilience to supply chain disruptions: Scale development and empirical examination. Journal of Operations Management, 33, pp. 111-122.

Bate, S. P., & Robert, G. (2002). Knowledge management and communities of practice in the private sector: lessons for modernizing the National Health Service in England and Wales. Public Administration, 80(4), pp. 643-663.

Bates, R., & Khasawneh, S. (2005). Organizational learning culture, learning transfer climate and perceived innovation in Jordanian organizations. International Journal of Training

and Development, 9(2), pp. 96-109.

Bessant, J., Kaplinsky, R., & Lamming, R. (2003). Putting supply chain learning into practice. International Journal of Operations & Production Management, 23(2), pp. 167-184.

Bhamra, R., Dani, S., & Burnard, K. (2011). Resilience: the concept, a literature review and future directions. International Journal Of Production Research, 49(18), pp. 5375-5393.

Bingham, C., & Davis, J. (2012). Learning Sequences: Their Existence, Effect, and Evolution. Academy Of Management Journal, 55(3), pp. 611-641.

Blackhurst, J., Dunn, K. S., & Craighead, C. W. (2011). An empirically derived framework of global supply resiliency. Journal of Business Logistics, 32(4), pp. 374-391.

Borins, S. (2001). Encouraging innovation in the public sector. Journal of intellectual capital, 2(3), pp. 310-319.

Botcheva, L., White, C. R., & Huffman, L. C. (2002). Learning culture and outcomes measurement practices in community agencies. American Journal of

Evaluation, 23(4), pp. 421-434.

Brandon‐Jones, E., Squire, B., Autry, C., & Petersen, K. J. (2014). A Contingent