key: cord-1019580-8izzhas3
authors: Gultekin, Beyza; Demir, Sercan; Akif Gunduz, Mehmet; Cura, Fatih; Ozer, Leyla
title: The logistics service providers during the COVID-19 pandemic: The prominence and the cause-effect structure of uncertainties and risks
date: 2022-01-13
journal: Comput Ind Eng
DOI: 10.1016/j.cie.2022.107950
sha: 3ff5852c0bde400657dd5b4bdfae7477da937bf5
doc_id: 1019580
cord_uid: 8izzhas3

Uncertainties and risks play a central role in creating vulnerabilities for logistics service operations. Over the years, Logistic Service Providers (LSPs) have learned how to ensure resilience to confront uncertainties and risks triggered by adverse events. However, quite unlike any seen in recent times, the COVID-19 pandemic brings about unavoidable uncertainties and risks for the logistics industry. Yet, there is no common approach to contextualize how they interact together. We incorporate an empirical research design and make a threefold contribution: first, we identify uncertainties and risks that LSPs encounter during the COVID-19 pandemic and investigate their prominence. Second, we unveil intertwined schemes of afore-identified uncertainties and risks and augment the understanding of their cause-effect structure. Third, we provide an uncertainty and risk assessment guideline for LSPs affected by threats emerging from unforeseeable crises. In this study, we combine qualitative work and the fuzzy DEMATEL method. Qualitative thematic analysis of in-depth interviews reveals the most important uncertainties (COVID-19 measures, employee welfare, forecast horizon, demand change, and government regulations) and risks (COVID-19 risk, delivery delays, supply chain disruptions, financial failure, and product returns) for LSPs. The fuzzy DEMATEL method shows that COVID-19 measures and COVID-19 risk are highly prominent and influence other factors. The results indicate that demand change, government regulations, and supply chain disruptions are net causers, and employee welfare, financial failure, forecast horizon, delivery delays, and product returns are net receivers. Distinctly, employee welfare is the most affected factor, empirically confirming that major risks for LSPs are related to the human factor. More investigation in our results suggests that supply chain disruptions and demand change, two factors triggered by the COVID-19 pandemic, influence financial failure and forecast horizon, two factors associated with operational performance.

The COVID-19 pandemic has wreaked devastation on global social and economic systems. The ramifications of this pandemic, or so-called great shock, have impacted manufacturing processes and worldwide supply chains (Chen, 2020) . Globalization accelerates this process of supply chain interdependence and creates difficulties for both individuals and the global supply chain, resulting in many uncertainties and risks on a global scale. Uncertainty and risk are inherent in all economic activities, albeit to varying degrees (Toma et al., 2012) . As supply chains become complex, uncertainties and risks become more significant (Choi, 2021; Shahbaz et al., 2019) . Turkey's exports fell by 41% in April 2020 due to the inability of LSPs to transport white goods, textiles, cars, and auto parts (Pitel, 2020) . The US automakers could not receive parts from Chihuahua (a Mexican state) plants because more than half of the workers were absent (Okamoto, 2020) . Supply chains are exposed to unprecedented supply chain disruptions (Ivanov and Das, 2020) , uncertainties (P. , and risks (R. during the COVID-19 pandemic. Uncertainty and risk are important research topics, and many papers investigate these concepts in the literature (e.g., Simangunsong et al., 2012; Fan and Stevenson, 2018; Prakash et al., 2017) . However, there is no coherence in the literature about how uncertainty and risk are conceptualized. Even though numerous studies have been conducted on the classifications of uncertainty (Sanchez-Rodrigues et al., 2008) and risk (Tarei et al., 2018) , there is a dearth of empirical research examining multiple types of uncertainties and risks simultaneously (Choi et al., 2019) . Even though uncertainty consists of multidimensions, it is usually considered a single dimension, namely the environmental dimension (Yu et al., 2017) , resulting in a poor understanding of the environment (P. . Nonetheless, a limited number of studies on various types of uncertainties and risks in conjunction with cause-effect interactions exist in the literature. In business activities, situations that are overlooked or not initially considered may arise, and these unforeseen events can create uncertainty, which can be a source of risk (Toma et al., 2012) . Defining and classifying risks enable companies to proactively manage uncertainties that may arise in the future (Hallikas et al., 2004) . There is a symbiotic relationship between uncertainties and risks, a field that requires additional research. There is a need to clarify the prominence and cause-effect relationships of uncertainties and risks that significantly impact the logistics industry. Prioritizing the uncertainties and risks and structuring their cause-effect relationships will contribute to understanding the business environment and enable companies to effectively allocate essential resources, develop employee welfare strategies to avoid workflow disruptions, ensure flexibility, and reorganize distribution against demand fluctuations and supply chain disruptions. The purpose of this study is to identify uncertainties and risks that LSPs encounter during the COVID-19 pandemic and investigate their prominence and cause-effect structure. This study reveals uncertainties and risks in the logistics industry by conducting in-depth interviews and thematic analysis and explores their cause-effect structure by employing the fuzzy DEMATEL (Decision Making Trial and Evaluation Laboratory) method. Figure 1 summarizes the basis of this research work. The solid lines in the model represent one-way causal effect relationships, while the dotted line symbolizes interdependencies among uncertainties and risks.

Types of Risks

Causal effect Interdependency Figure 1 . The components under examination This paper is organized as follows. Section 2 presents the literature for uncertainty and risk. Section 3 describes the methodological framework of in-depth interviews, thematic analysis, and the fuzzy DEMATEL method and its corresponding algorithm. Section 4 exhibits the results and the sensitivity analysis. Section 5 addresses the theoretical and managerial implications. Sections 6 and 7 demonstrate conclusions and discussion and limitations and directions for future research, respectively.

While more recent attention focuses on the uncertainty and risk subjects in several research domains, a large and growing body of literature investigates the uncertainty and risk issues concerning the logistics industry. Various studies (Prakash et al., 2017; Wang et al., 2019 Wang et al., , 2014 consider uncertainty and risk as similar concepts due to their being indivisible and their simultaneous management in practice (Wang, 2018) , whereas other studies (Sachs, 2018; Shahbaz et al., 2019; Simangunsong et al., 2012; Stewart, 2021 ) describe uncertainty and risk as different concepts. Essentially, the uncertainty may have negative and positive consequences, whereas the risk is solely related to negative results 4 (Simangunsong et al., 2012) . This section discusses the literature of uncertainty and risk concepts and reveals the types of uncertainties and risks. The existing literature particularly focuses on the multidimensionality of uncertainties and risks (Stewart, 2021) . More frequently, studies concentrating on uncertainty or risk regarding the logistics industry adopt structural equation modeling or a case study approach. A few studies investigate relationships among logistics risks or uncertainties; however, none examine these two subjects simultaneously. The nature of how uncertainties and risks interact concerning the logistics service domain remains unclear. Table 1 compares similar studies and addresses the literature gap. Bae (2012) X Support-vector machines Logistics industry Lin et al. (2013) X Principle-agent modeling Logistics service industry Kazemi Zanjani and Nourelfath (2014) X Stochastic modeling Service industry X Quality control game model Logistics service industry X Scheduling model Logistics service industry Govindan and Chaudhuri (2016) X DEMATEL Logistics service industry Moslemi et al. (2016) X Case study Logistics industry Baharmand et al. (2017) X After briefly reviewing the literature on uncertainty and risk concepts, this section outlines the several types of uncertainties and risks. Then, the prominence of uncertainties and risks and their cause-and-effect structure are examined.

Uncertainty exists when there are several possible outcomes, some or all of which are unknown, and the probability of each cannot be calculated (Stewart, 2021) . Uncertainty arises when it is impossible to predict an outcome's likelihood and the consequences of a decision (Sanchez-Rodrigues et al., 2010b) . Uncertainty stems from the variability resulting from the changes in nature, social environment, and technology and limited knowledge due to the lack of observations and measurements and unreachable information (Wattanakul et al., 2019) . Uncertainty limits managers' decision-making capabilities by preventing them from determining the variances and probability of occurrences (Christopher and Lee, 2004; Wang, 2018) . However, this does not mean that uncertainty is always detrimental to businesses; it is associated with positive and negative outcomes (Simangunsong et al., 2012) . For instance, profit increases or decreases when demand uncertainty exceeds or falls short of expectations (Simangunsong et al., 2012) . Uncertainty may stem from a variety of sources. For example, uncertainty is caused by the supplier, carrier, customer, control systems, and other external variables (Sanchez-Rodrigues et al., 2008) . In other words, uncertainty can be traced throughout the supply chain, whether on the ship, in the port, or during transit (Wattanakul et al., 2019) . We classify LSPs' uncertainties into four categories: supply, demand, internal, and external uncertainty (see Table 2 ). Supply uncertainty includes uncertainties related to the forecast horizon, shipper, provider, production, and supply chain process. Demand uncertainty is related to the customer or receiver of the products. Internal uncertainties are related to the process and the operations and include inefficiencies originated by the carrier. External uncertainties arise due to the environment, government regulations, competitor behavior, macroeconomic concerns, natural disasters, and pandemics such as COVID-19. 

Risk is an event's probability of occurrence and adverse consequences (Calatayud et al., 2017) . The greater the impact of the threat on the supply chain, the greater the risk is, even if it is less likely to occur (Sanchez-Rodrigues et al., 2010b) . The risk is a threat that interrupts regular activities and end planned operations which may negatively affect credibility, reputation, and trust (Fan and Stevenson, 2018) . There are various types of risk classification in the supply chain management domain, and more specifically in the logistics sector. Types of risks in a supply chain are logistics, information, and financial risks (Shahbaz et al., 2019) , customer-side, company-side, and environment-side . According to Baharmand et al. (2017) , logistics risks include delivery delays, market fluctuations, insufficient capacity, cargo loss and decay, unreliable information, and ethical problems. We classify LSPs' risks into four categories: supply, demand, internal, and external risk (see Table 3 ). Supply risks are related to supplier performance (e.g., supply chain disruptions, the bankruptcy of suppliers, delays in supply lead-time, short supplies, outsourcing, and control), products (e.g., poor quality of supplies), or process (e.g., disruption in production, low production capability, inflexibility in capacity, low production yield, wrong order quantities, process risks related to value-added activities, property damage, low capacity). Demand risks are customer-related issues such as fluctuations in demand, seasonality, volatile customers, change in customer preference, inaccurate forecasting of demand, payment failure, and fraud. Internal risks are financial failure (e.g., cash flow issues, intellectual property, merger/alliance, and trust), delivery delays, leadership and management style (e.g., ethical concerns, market perceptions), collaborations, buyer and supplier relationships, sustainability, information flow, IT system failure, loss or damage of cargo, unavailability of necessary vehicles, improper loading, and mode of transport errors. External risks include force majeure such as COVID-19 pandemic, natural and man-made disasters, market fluctuations, economic crisis, and political, social, and industrial issues. 

The prominence of uncertainties and risks is vital for the allocation of resources. According to Nguyen et al. (2021) , physical flow is the primary source of high-ranking risks with potentially serious consequences (e.g., piracy, dangerous cargo, and maritime accidents), whereas informational, financial, and operational issues (e.g., fuel costs) are more uncertain. Uncertainty generates (Choi et al., 2019) and amplifies (Christopher and Lee, 2004) risk, increases the probability of risk (Wang, 2018) , and results in a more complex supply chain (Christopher and Lee, 2004) . For instance, supply uncertainty arising from contractual obligations regarding volume or mix delays the scheduled delivery and increases the delivery risk (Sreedevi and Saranga, 2017) . The uncertainties related to inventory holding costs and shipment frequency create more risk than volume changes and freight rates (Christopher and Lee, 2004) . Supply and demand-side uncertainties can also create supply chain disruptions and logistics risks (Choi et al., 2019) . More specifically, road network congestion, an external uncertainty, is one of the primary reasons for the delivery delays, an internal risk. Volatile demand, a demand risk, occurs due to unexpected promotions (Sanchez-Rodrigues et al., 2010a) , an internal uncertainty. Uncertainty becomes a cause of risk when it is based on incomplete information or derived from sources that are frequently incompatible with the actual state of a business or competitive market (Toma et al., 2012) . On the contrary to popular belief, risks are not only stem from uncertainties; simultaneously, they increase the possibility of uncertainty in the supply chains (Sanchez-Rodrigues et al., 2010a) . Risks contribute to uncertainty in a complex environment such as logistics (Calatayud et al., 2017) .

Internal risks such as ineffective information flow or coordination between marketing and logistics departments create uncertainty (Sanchez-Rodrigues et al., 2010a) . Global supply chain risks (e.g., inadequate departmental coordination) and supply and demand risks (e.g., mismatching supply and demand with insufficient buffer stock) (Stewart, 2021) may cause uncertainty in the supply chain. A strong interdependence exists between uncertainty and risk (Jedynak and Bąk, 2020) , necessitating a collaborative action (Hallikas et al., 2004) . Systems theory conceptualizes an organization as a network of interconnected systems, a "system of systems" (SoS), which guides in understanding the risks as constitutionally intertwined (Fan and Stevenson, 2018) . For example, supply disruptions, a supply risk, and operational problems in unloading and loading, an internal risk, results in delivery delays, another internal risk; volatile demand, a demand risk, may occur due to, and changes in customer preferences, another demand risk (Sanchez-Rodrigues et al., 2010a) . Along with risks, uncertainties are a critical part of this SoS approach (Hou and Zhao, 2020) . For instance, supply uncertainty is related to demand uncertainty. Thus, one can argue that uncertainties are intertwined like risks. The SoS approach regarding uncertainty also urges the LSP managers to set strategies to mitigate risk (Hou and Zhao, 2020) . Risk management in logistics is challenging because of the complicated and fragile environment shaped by uncertainties (Shahbaz et al., 2019) . In this sense, managers can take additional precautions when they are aware of these intertwined relationships rather than treating uncertainties and risks independently. Risks and uncertainties identified, evaluated, and compared may require joint action or be incorporated into planning processes (Hallikas et al., 2004) . However, LSPs are still unclear about the types of uncertainties and risks and their interaction during the COVID-19 pandemic. Thus, this symbiotic relationship among uncertainties and risks necessitates further research to understand their underlying cause-effect structure.

Understanding the prominence of uncertainties and risks and their causal relationship enables LSPs to diagnose their vulnerabilities. Accordingly, this study proposes three research questions: RQ1. What are the uncertainties and risks encountered by the LSPs during the COVID-19 pandemic? RQ2. How prominent the determined uncertainties and risks are for LSPs during the COVID-19 pandemic? RQ3. What is the cause-effect structure of these uncertainties and risks?

This study's methodology introduces a literature review to identify the relevant uncertainties and risks encountered by the LSPs during the COVID-19 pandemic. In-depth interviews with experts follow this to specify primary uncertainties and risks. Then we analyze the data collected from managers of LSPs using a fuzzy DEMATEL method to explain the complex causal structure and interdependencies among these uncertainties and risks. Figure 2 illustrates the research's flowchart to analyze and prioritize the significant uncertainties and risks for LSPs during the COVID-19 pandemic. The research plan consists of six stages: (1) identifying the uncertainties and risks by reviewing the literature and conducting in-depth interviews with experts and specifying the uncertainties and risks and the questionnaire design, (2) linguistic data collection via conducting the questionnaire with managers of LSPs and construct results by triangular fuzzy numbers, (3) defuzzificating triangular fuzzy numbers by the CFCS method, (4) data analysis by employing the fuzzy DEMATEL method, (5) conducting the sensitivity analysis to test the robustness of the managers' evaluations, and (6) presenting the overall prominence and causal effect diagram and the conclusion.

Step 2. Conduct indepth interviews with experts

Step 6. Determine the pairwise interactions among uncertainty and risk elements

Step 7. Transform the linguistic expressions into triangular fuzzy numbers

Step 8. Standardize the fuzzy numbers

Step 9. Obtain the crisp numbers for each manager's assessments

Step 10. Aggregate the crisp scores for all managers

Step 11. Compute the normalized initial direct-relation matrix

Step 12. Compute the total relation matrix

Step 13. Calculate the total and net effect values

Step 14. Conduct the sensitivity analysis to test the robustness of the results

Step 15. Extract the cause-effect model of uncertainties and risks

Step 3. Analyze qualitative data by thematic coding

Step 4. Identify the uncertainties and risks according to the findings of qualitative analysis

Step 1. Review the literature on uncertainties and risk affecting the vulnerability of LSPs

Step 5. Acquire managers' linguistic evaluations via questionnaire 

Since the uncertainty and risk elements presented in Section 2 are based on the general logistics service domain, it is not valid to make direct inferences for the specific conditions of the COVID-19 pandemic. However, the uncertainty and risk elements identified by reviewing the literature provide a starting point for the following in-depth interviews used to identify uncertainty and risk elements typical to the logistics service industry conditions shaped by the COVID-19 pandemic. Qualitative research is preferred when the research subject is relatively unexplored since it is more eligible to elicit new information and support phenomenological validity (Vanderstoep and Johnson, 2009 ). In particular, in-depth interviews, a qualitative data collection technique, allow conceptualizing a phenomenon by capturing contextual information (Trochim and Donnelly, 2006) . Following the literature review, we interviewed experts to capture their opinions to identify the most significant uncertainty and risk factors for the LSPs working in the COVID-19 environment. We employ in-depth interviews based on the procedure used by Verkooij and Spruit (2013) . The expert selection protocol is based upon two criteria. First, experts should have broad experience and practical knowledge in the logistics services industry. Second, they should be actively involved in logistics service operations or research. Having this up-to-date knowledge and experience would allow experts to make sound judgments on the uncertainties and risks that affect the vulnerability of LSPs during the COVID-19 pandemic. Accordingly, we determine a sample of three academics and three executives with over fifteen years of experience in logistics services. Table 4 lists the six experts interviewed in this study who are adhered to the defined selection criteria. During the in-depth interviews, the interviewees are asked to clarify the most significant drivers affecting LSPs' operations within the scope of the COVID-19 pandemic. Follow-up questions explore the predictability and eventuality of these drivers to support classifying themes as uncertainty or risk. The in-depth interviews are tape-recorded and transcribed. The transcript summaries are verified by the interviewees. Then, the interviews are analyzed using thematic coding and compared to and associated with the reviewed uncertainty and risk typology by grouping closely related narratives concerning the uncertainty and risk themes. The thematic coding is performed by adapting the conceptual framework proposed by Flick (2018) . A three-digit coding procedure is followed for the coding process of the qualitative data. Apriori codes are employed to reflect categories complying with the literature review (Gibson and Brown, 2009) . As reviewed and classified in Section 2, the literature on uncertainty emphasizes supply, demand, internal, and external uncertainties, and the risk literature focuses on supply, demand, internal, and external risks. The first digit of the code classifies a theme as uncertainty (U) or risk (R), the second digit signifies uncertainty or risk subcategory (S=Supply, D=Demand, I=Internal, E=External), and the third digit designates the element number (e.g., 1, 2, 3, 4). The thematic coding process is held in two stages. At the initial scanning stage, twelve uncertainty themes and twenty-four risk themes are identified. At the elimination stage, these are reduced to five for each theme. As a threshold procedure, a minimum of three experts' (half of the sample group) narratives is needed to refer to reasonable consideration of defining a theme within the uncertainty and risk framework. The rationale for this approach is that the themes would exclude highly specific narratives but keep enough details to distinguish various uncertainty and risk elements. Table 5 exhibits the identified uncertainty and risk themes at the initial scanning stage and the qualitative analysis coding system. Three executives expressed concern for employee welfare during the interviews, while three interviewees (one professor and two executives) expressed concern about the product returns. Employee welfare is an underestimated area of research, and there are limited studies that consider employees as a part of uncertainty and risk and do not separate uncertainty from risk. For example, Wang et al. (2015 Wang et al. ( , 2018 identify inadequate communication between the company and its drivers as an internal uncertainty and risk, while labor/driver shortage is classified as an environmental uncertainty and risk. Sanchez-Rodrigues et al. (2010b) consider employee-related issues under uncertainty and state how driver inadequacy contributes to the internal uncertainty related to the carriers. In this perspective, businesses view uncertainties and risks associated with employees primarily through the lens of inefficiency, communication, and scarcity. In this study, employee welfare is defined as an internal uncertainty since it cannot be predicted and is expected to have positive or negative results. There is no mention of product returns in the literature on uncertainty and risk. However, the product returns literature acknowledges returns as a risk for businesses (Padmanabhan and Png, 1997) . Padmanabhan and Png (1997) consider returns policies a risksharing mechanism in the business-to-business (B2B). Fu et al. (2016) identify an imbalanced distribution of consumers' product-return biases and develop a model for identifying the riskiest product returns (i.e., most probable to return or not). Rao et al. (2014) analyze the impact of physical distribution services in product returns and find that managing expected delivery timeliness and delivery reliability and maintaining delivery promises reduces the risk of product returns. Martino et al. (2015) mention the product return risk related to the quality check in the central warehouse or logistic center and post-season unsold goods. According to the interviews and literature, this study considers product returns to be a demand risk. Thus, examining employee welfare and product returns in the context of LSPs is a rare field of research that contributes to the literature on uncertainty and risk. We identified five uncertainty and five risk elements through the elimination stage of thematic analysis. Uncertainty factors are forecast horizon (a supply uncertainty), demand change (a demand uncertainty), employee welfare (an internal uncertainty), government regulations (an external uncertainty), and COVID-19 measures (an external uncertainty), and risk factors are supply chain disruptions (a supply risk), product returns (a demand risk), financial failure (an internal risk), delivery delays (an internal risk), and COVID-19 risk (an external risk). Table 6 presents the items retained after the thematic analysis of the qualitative data. Accordingly, we reorganize identified uncertainties and risks and conceptualize them in conformity with the literature. 

Following the expert interviews, we surveyed managers of fifteen LSPs operating in Turkey to elicit the cause-effect relations and interdependencies among uncertainties and risks. For data collection, the expert sampling method is employed, which is a subcategory of purposive sampling. Purposive sampling involves selecting the sampling units related to the most information on the specific subject (Guarte and Barrios, 2006) . Expert sampling suggests collecting data from a sample of people known for their experience and expertise in the field (Trochim and Donnelly, 2006) . Firstly, we identify companies engaged in activities in the logistics service industry for more than five years. Then, we contact the managers of these LSP companies, who have experience in logistics service operations for more than five years, to capture their opinions on the uncertainties and risks for LSPs during the COVID-19 pandemic. A questionnaire is used to obtain the managers' evaluations on the pairwise interactions among uncertainty and risk elements specified by the in-depth interviews.

Respondents are also asked about the company's activity branch, age, size, position, and experience. It takes from 45 to 60 minutes for one respondent to complete the questionnaire. Table 7 lists the fifteen managers surveyed in this study. The questionnaire is composed of four parts. The first part describes each uncertainty and risk to make it clear for responding. In the second part, respondents are asked to rate the significance of each uncertainty and risk using a five-point Likert scale. The ratings of "1, 2, 3, 4, and 5" represent "not at all important", "slightly important", "moderately important", "very important", and "extremely important", respectively. The third part is a five-point Likert scale pairwise association matrix to assess the influence of uncertainties and risks on each other. The ratings of "0, 1, 2, 3, and 4" represent "no influence", "very low influence", "low influence", "high influence", and "very high influence", respectively. Finally, the fourth part covers organizational demographics.

In this study, we use the fuzzy DEMATEL method to determine the prominent uncertainty and risk factors that affect the vulnerability of LSPs under overwhelming COVID-19 conditions and evaluate their cause-effect structure. We consider these factors as interrelated elements, reflecting that various uncertainties and risks influence each other. The rationale for selecting the DEMATEL method is its methodological advantage over other decision-making methods to handle complex causal relationships and interdependencies (Si et al., 2018) . The DEMATEL method aims to find integrated solutions for fragmented and antagonistic phenomena (Wu, 2008) . It is an enhanced procedure for analyzing and designing a structural model to measure complicated causal relations among multiple criteria (Chang et al., 2011) . It features complex cause-effect relationships between a set of elements through matrices (Wu and Lee, 2007) . It has a methodological capability to analyze the pairwise interrelations between items (Si et al., 2018) . The DEMATEL method presumes that all factors have different levels of effects on each other, i.e., they are not independent (Tzeng and Shen, 2017) . Compared with other decision-making methods, the DEMATEL method has the following advantages (Lin and Tzeng, 2009; Liu et al., 2014; Si et al., 2018; Tzeng et al., 2007) : (1) it can effectively analyze the direct and indirect effects among different factors and enable the decision-maker to understand the complex causal structure, (2) it can visualize the interrelationships between factors via a cause-effect model, allowing the decision-maker to understand which factors have influences on the others, (3) it can determine the prominence of factors. In recent years, considering the ambiguity of respondents' opinions, many scholars (e.g., Ocampo et al., 2019; Sathyan et al., 2020; Zhang and Su, 2019; Zhou et al., 2018) combine fuzzy logic techniques with the DEMATEL method to satisfy a solution for the vagueness in these complex problems. We analyze the managers' data with the fuzzy DEMATEL method similar to previous studies (e.g., Addae et al., 2019; Chang et al., 2011; Sathyan et al., 2020; Zhou et al., 2011) . The steps of the fuzzy DEMATEL method are explained in the next sections.

In the first phase, we develop a direct-relation matrix for each of fifteen managers using the linguistic scale, then set the corresponding fuzzy triangular numbers, and finally find fifteen direct-relation fuzzy matrices.

Step 1. Identifying the uncertainties and risks Firstly, we identify the uncertainties and risks as described in the questionnaire design section. The pairwise interactions between uncertainties and risks for LSPs come from managers.

Step 2. Assessing the pairwise interactions among uncertainty and risk elements The second step of the fuzzy DEMATEL method finds the direct-relation matrices = with the data collected from managers for uncertainties and [ ] × ( = 1, 2, …, ) risks . The diagonal elements of direct-relation matrices are equal to zero ( = 1,2,…, ) since an item cannot directly influence itself according to the DEMATEL method (Tzeng et al., 2010) .

Step 3. Transform the linguistic expressions into triangular fuzzy numbers Importance weights and influence scores designated by the managers are essentially linguistic expressions. As Li (1999) suggests, for tackling the vagueness of managers' assessments, the ratings of the linguistic influence scale are converted into triangular fuzzy numbers (see Table 8 ). Let denotes the rating of uncertainty/risk that influences = ( , , ) uncertainty/risk , obtained from managers. In the fuzzy transformation of the lingual expression, signifies the smallest possible value, indicates the most probable value, and represents the largest possible value of the fuzzy phenomenon. of managers for each uncertainty and risk pair. According to the fuzzy transformation, each manager response corresponds to a fuzzy number to define the direct relations between uncertainty/risk pairs. For example, with uncertainties/risks and managers ( = 1,2,…, ) , the direct-relation fuzzy matrices can be built as below where ( = 1, 2, …, ) = :

In the second phase, we defuzzy the direct relation fuzzy matrices by the "Converting Fuzzy data into Crisp Scores" (CFCS) defuzzification method provided by Opricovic and Tzeng (2003) .

Step 4. Standardize the fuzzy numbers According to the CFCS defuzzification method, the standardized fuzzy values are calculated as "a weighted average according to the membership functions" by using Equations (1)-(3):

Where , Δ = max -min

Step 5. Obtain the crisp numbers Next, we defuzzy the values of the triangular fuzzy number matrices to obtain the crisp values , and find the direct-relation defuzzy matrices . The crisp numbers are

calculated according to the CFCS method by using Equation (4):

Where ;

Thus, we transform fuzzy numbers into crisp values.

Step 6. Aggregate the crisp scores In the sixth step, the average values of the direct-relation defuzzied matrices are calculated by dividing the sum of causal effect ratings by the number of respondent managers (15) and the average direct-relation matrix is obtained (see Appendix I).

We calculate the average direct relations for all managers by using Equation (5) 

In the third phase, the cause-effect relationships among uncertainties and risks are determined, and their prominence is revealed according to the analysis results.

Step 7. Compute the normalized initial direct-relation matrix Here, we normalize the average initial direct-relation matrix W by using Equation (6) Step 8. Compute the total relation matrix The total relation matrix is set up from the total effects that uncertainty/risk

gives and receives by using Equation (7) (see Appendix III) . The values in the total relation matrix represent the sum of the row factors' direct and indirect influence on the column factors (Hinduja and Pandey, 2018) . A row factor's direct influence on a column factor indicates an unmediated causal effect, and indirect influence signifies a mediated causal effect through the system (Tzeng and Shen, 2017) .

Step 9. Calculate the total and net effect values In the final step, firstly, we calculate the influential ( ) and influenced ( ) effects and the total ( ) and net effect ( ) values. Then we set a threshold value and acquire the + interdependency matrix and the cause-effect model to construct and visualize the causeeffect relationships among uncertainties and risks. We calculate the sum of rows (causal effects given by uncertainty/risk ) and columns (causal effects received uncertainty/risk ) of total relation matrix and find the influential effect and influenced effect for each uncertainty and risk by using Equations (8)

The influential effect values indicate the sum of direct and indirect causal effects given by uncertainty/risk to other uncertainty and risk elements in the system, the influenced effect values denote the sum of direct and indirect causal effects received by uncertainty/risk from uncertainty and risk elements in the system (Tzeng and Shen, 2017) . Then we calculate the total effect values ( ) and the net effect values ( ) + for each uncertainty and risk. According to Tzeng and Shen (2017) , the total effect value represents the sum of effects given and received by uncertainty/risk , while the net effect value designates the degree of net causal effect that uncertainty/risk has on the system. Uncertainty/risk has a positive net effect value if its total causal effect ( -) > 0 given on the system is higher than its total causal effect received from the system. Uncertainty/risk has a negative net effect value if its total causal effect ( -) < 0 received from the system is higher than its total causal effect given to the system. The total effect value is the degree of how prominent uncertainty/risk are. And the net effect value is the extent to which uncertainty/risk can affect the system and stir up other uncertainties and risks.

In this section, the analysis results of the fuzzy DEMATEL method are presented. The influential and influenced effects and the total and net effect

values of each uncertainty and risk are shown in Appendix IV. The prominence of uncertainties and risks are prioritized based on the total effect values. Figure 3 presents the prominence of the uncertainties and risks arranged in the descending order of total effect value. The COVID-induced uncertainties and risks top the system. Followingly, forecast horizon, financial failure, delivery delays, supply chain disruptions, and demand change are moderately prominent. Government regulations, employee welfare, and product returns are the least prominent factors. effects, which means COVID-19 measures and COVID-19 risk are the system's main drivers, i.e., net causers. Besides, government regulations, demand change, and supply chain disruptions are also grouped as net causers. Employee welfare has the lowest net effect indicating the system's main outcome, i.e., net receiver. Four other uncertainty and risk elements are identified as net receivers: forecast horizon, product returns, delivery delays, and financial failure. The managers are asked to evaluate each uncertainty and risk's significance within the questionnaire besides assessing interrelations. The significance criteria (see Table 9 ) indicate the average of fifteen managers' evaluation scores on the importance of each uncertainty and risk for their operations. The managers perceive COVID-19 risk, employee welfare, and COVID-19 measures as more significant, with the highest absolute net effect values. Nonetheless, the least significant factor is product returns, followed by government regulations, financial failure, and delivery delays. The remaining factors, forecast horizon, supply chain disruptions, and demand change, have moderate significance. Figure 5 shows the overall prominence and causal effect diagram based on the total and net effect values and the significance criteria. The overall prominence and causal effect diagram is formed by the horizontal axis that shows the total effect and the vertical axis that ( + )

shows the net effect values of uncertainties and risks. The horizontal axis shows ( -) how significant uncertainty/risk is, while the vertical axis classifies uncertainties and risks into net causer and net receiver groups. The upper part of the diagram contains net causers, while the lowermost comprises net receivers. The diameter of the circles indicates the significance criteria in Table 9 . We take the fourth power of the significance criteria to make it more distinctive in the diagram. We develop a cause-effect model (see Figure 6 ) to visualize the complex causal relationships and interdependencies among uncertainties and risks using the values of the interdependency matrix (see Appendix V) and total and net effect values. The

interdependency matrix shows the pairwise causal effect values obtained by removing elements less than the threshold value in the total relation matrix . We set a threshold value to filter out some negligible effects in the total relation matrix to explain the structural relations among the uncertainty and risk factors while keeping the system's complexity manageable (Tzeng and Shen, 2017) . Only uncertainty and risk factors in the total relation matrix having an effect greater than the threshold value are kept in the interdependency matrix (Tzeng et al., 2007) . In the literature, the threshold value is usually determined by decision-maker or expert discussions (Lin and Tzeng, 2009) , averaging the values of the total relation matrix (Quezada et al., 2018) , or taking the maximum value of the diagonal elements of the total relation matrix (Tan and Kuo, 2014) . If the threshold value is too low, the cause-effect model will be too complex for decisionmaking; in contrast, if the threshold value is too high, too many factors will be shown as independent elements without interacting with other factors (Tzeng et al., 2007) . This study sets the threshold value as the maximum to keep every element interacting with ( = 0.42) the system while reducing the complexity of the cause-effect model. Figure 6 shows the solution for the model in Figure 1 suggested in the Introduction Section. The extracted solution in Figure 6 illustrates the cause-effect model of uncertainties and risks based on the interdependency matrix in . The solid lines represent one-way causal relationships, while dotted lines symbolize interdependencies. The lines are colored as shown in the legend to disambiguate cause-effect relationships and interdependencies among uncertainties and risks. The results show that two uncertainties affect two risks, two risks affect three uncertainties, and one uncertainty and one risk are interdependent. Plus, three uncertainties affect two other uncertainties, and two risks affect four other risks. Specifically, COVID-19 risk, a net causer, affects all net receivers, i.e., employee welfare, forecast horizon, delivery delays, financial failure, and product returns, plus two net causer group elements, demand change, and government regulations. Two net causers, COVID-19 measures and demand change, affect all net receivers, but product returns which is only affected by COVID-19 risk. Forecast horizon, another net receiver, is affected by all five net causers, i.e., demand change, government regulations, COVID-19 measures, supply chain disruptions, and COVID-19 risk. Two other net receivers, delivery delays and financial failure, are affected by the same four net causers, all net causer group elements but government regulations. Another net receiver, employee welfare, is also affected by four net causers, but by government regulations instead of supply chain disruptions. Only interrelation in the extracted model is between two net causers, COVID-19 measures and COVID-19 risk. As a result, the proposed two-way causeeffect relationship and interdependency between uncertainty and risk types are confirmed.

Decision-makers' risk sensitivity influences their perception of risk and uncertainty magnitude (Tversky and Fox, 1995) . Moreover, the actions of risk-prone or risk-averse decision-makers are extensively associated with environmental uncertainty and risk (Ben-Haim, 2000) . Hence, managers' perceptions should be addressed to validate the results since risk-sensitivity influences the decision-making behavior in supply chain operations (Tsay, 2002) . Figner and Weber (2011) suggest that risk-taking behavior depends on situational and idiosyncratic characteristics. Therefore, to estimate the degree of risk sensitivity in the robustness test, managers' evaluations are weighted according to job experience, management level, job responsibility, and average risk/uncertainty importance score. The robustness of the analysis results is confirmed by conducting the sensitivity analysis to test the dependability of the managers' evaluations. In line with the studies utilizing sensitivity analysis to test the DEMATEL method's robustness (Bhatia and Srivastava, 2018; Govindan et al., 2015; Seker and Zavadskas, 2017) , we outline scenarios applying various combinations of different values of manager attributes. Initially, in Scenario A, equal weights are assigned to each manager. Subsequently, in Scenarios B to E, each manager's weights are altered in terms of manager attributes to analyze the causal effect relationships' variation. The weights of each manager vary in designated scenarios since they have different degrees of job experience (Scenario B), managerial level (Scenario C), job responsibility (Scenario D), and average uncertainty/risk importance score (Scenario E). The causal effect values obtained from different sensitivity analysis scenarios are presented in Figure 7 , drawing two lines for each scenario, one for the total causal effect with higher values and one for the net The sensitivity analysis suggests that the DEMATEL method's results are valid and not highly dependent on the number of participants. The structure of cause-effect relationships is consistent in different scenarios, indicating that the results reflect consulted managers' genuine opinions. To conclude, managers' responses on the causal structure of uncertainties and risks in the logistics service industry are sufficient for this study.

Several implications for LSP managers can be drawn from the proposed method and its relevant findings. Our study identifies the prominence and cause-effect structure of uncertainties and risks LSPs confront during the COVID-19 pandemic and guides logistics managers to determine businesses' vulnerabilities. Identifying LSPs' weaknesses and strengths is important not only in the period of COVID-19 but also in guiding them for any unexpected supply chain disruption in the future. In this way, LSPs will establish more flexible, agile, and resilient service systems by taking precautions against unexpected situations.

Our study focuses on discovering uncertainties and risks LSPs face during the COVID-19 period and analyzes their prominence and cause-effect structure. The proposed uncertainty and risk assessment framework offers managers and decision-makers a systematic approach to prioritize measures and decisions against COVID-19 induced risks and uncertainties. The framework augments understanding the relationship between risks and uncertainties by disclosing the cause-effect model (see Figure 6 ) that illustrates the interrelations between risks and uncertainties under consideration. Hence, LSPs operating in volatile markets will be able to build flexible, agile, and resilient business models against unforeseen risks and uncertainties. The following paragraphs in this section describe the managerial implications. Employee welfare, forecast horizon, financial failure, delivery delays, and product returns form the net receiver group. Employee welfare is the most affected factor since it has the lowest net causal effect value, implying that it is the most influenced component of the system. Employee welfare is particularly influenced by COVID-19 risk, COVID-19 measures, government regulations, and demand change, respectively. This finding is in line with the study of Dorofeev et al. (2020) that emphasizes that shipping companies' major risks are associated with human resources. Okamoto (2020) mentions the health of the employee as a recent operational risk. Although heavy truck drivers are excluded from many of the restrictions, they are unwilling to come to work due to the fear of becoming infected, and even there are cases so and more severe than that (Dorofeev et al., 2020) . Besides, the supply problems in materials that need labor intensity will form a considerable share of supply chain disruptions (Chenneveau et al., 2020) . However, employee welfare is an underestimated area of research. McKinsey & Company (2021) emphasizes the importance of providing a guideline when dealing with COVID-19, granting autonomy to the employees in a rapidly encountered situation, and creating a two-way communication style in terms of feeling safe. Systems to support remote working conditions and contactless logistics services (touch-free payment, mobile robots, delivery by drones) also contribute to employee welfare. A few studies (Sanchez-Rodrigues et al., 2010b; Wang et al., 2015 Wang et al., , 2018 consider employees a part of risk and uncertainty. Sanchez-Rodrigues et al. (2010b) highlight that driver insufficiency leads to uncertainty, which is among the inefficiencies originated by the carrier. In this sense, companies consider uncertainties and risks sourcing from the employees solely in communication, shortage, and inefficiency. Wang et al. (2015) define "poor communication between company and drivers" as an internal uncertainty and risk and classify "labor/driver shortage" as an environmental uncertainty and risk. Since employee welfare relates to employee satisfaction, it is vital for companies (Bandara et al., 2020) . Forecast horizon and financial failure are among the most affected factors mainly influenced by COVID-19 measures, COVID-19 risk, government regulations, demand change, and supply chain disruptions. Volatility forecasts improve financial risk management (Christoffersen and Diebold, 2000) and help companies reduce the financial failure risk by predicting payment, cash flows, and delivery disruptions. Therefore, companies that can mitigate financial risks will be capable of making accurate forecasts. COVID-19 measures that include several arrangements (e.g., wearing masks, physical distancing, hygiene) impact the forecast horizon. COVID-19 risk has low predictability, thus complicating the long-term forecasting and unearths financial failure risks. Companies are urged to guarantee liquidity, make model simulations, and identify the factors that threaten their liquidity to prevent financial failure (McKinsey & Company, 2021) . Besides, ensuring options to hedge in logistics, e.g., shipping quantity, delivery date/time, price (Tibben-Lembke and Rogers, 2006) , pinpointing reliable suppliers, rebuilding a faster logistics system, and being less costly due to the advanced technology (Chenneveau et al., 2020) help companies make decisions wisely. Product returns and delivery delays are also important factors affected by the net causer group elements. Product returns, the least prominent factor in the system, are influenced by COVID-19 risk, COVID-19 measures, demand change, and supply chain disruptions. Efficient management of reverse logistics is a vital part of the supply chains (Potdar and Rogers, 2012) ; managers use forecasting methods related to product returns that provide cost savings for remanufacturing (Clottey et al., 2012) . Reverse logistics can be supported by authorizing third-party logistics companies to manage the maintenance, repair, and operations (MRO) function (Suyabatmaz et al., 2014) . Delivery delays are influenced by COVID-19 risk, COVID-19 measures, demand change, supply chain disruptions, and government regulations. This result is associated with lockdowns, restrictions, measures, demand/supply fluctuations due to the COVID-19 pandemic (Barua, 2020 (2021) reports that customer demand is the most influencing uncertainty for supply chain and production managers. We conclude that COVID-19 measures and COVID-19 risk take the first and second place in terms of the total impact on the LSPs. These are followed by government regulations, demand change, and supply chain disruptions, respectively. Governments are policymakers or influencers that affect the external environment within which logistics operations are held. The relation between efficiency and environmental impact becomes clear for logistics managers today. For instance, government interventions such as taxation laws or regulations can stimulate many businesses to change their core strategies (Sanchez-Rodrigues et al., 2010b) . Health-related government regulations such as lockdowns have an impact on supply chains. According to Mollenkopf et al. (2020) , sudden shifts in health-related regulations result in serious supply chain disruptions such as farmers not being available to harvest crops, shutdowns of the foodservice and restaurant sector, and productivity decline due to changing working conditions. China's social logistics costs have significantly increased during the COVID-19 period. Liu et al. (2020) state that some of the reasons for this rise are the worker shortage caused by the people's mobility restriction and struggles in planning transportation routes due to the uncertainties in traffic restrictions. The demand change and supply chain disruptions impose difficulties on the companies making decisions on investments, manufacturing, scheduling, and forecasting. Chenneveau et al. (2020) recommend coordinating the demand planners with the sales department and the data analysts to forecast demand accurately. Crawford (2020) suggests that coordinating the continuous evaluation of the effects of COVID-19 will enable the supply chain to operate more efficiently. Similarly, Jiang et al. (2020) demonstrate that emergency coordination and command systems support logistics reliability.

To survive in an uncertain and risky environment of COVID-19, it is recommended to diversify suppliers and omnichannel distribution (McKinsey & Company, 2021) , hold more inventory, invest in automation (Okamoto, 2020) , and reorganize the inventory management system (Crawford, 2020) . As company executives may face pandemic-like disruptions in the future, they should develop some digital strategies to overcome problems such as market uncertainty and supply chain challenges. These strategies include adapting to technologies such as the Internet of Things, artificial intelligence, robotics, and 5G, as the digital transformation of supply networks is geared to anticipate and solve future challenges with advanced features (Deloitte, 2020) . Moreover, adopting Industry 4.0 technologies such as horizontal and vertical integration, augmented reality, cloud computing, blockchain technology will help companies reduce COVID-related uncertainties and risks by decreasing human intervention.

In-depth interviews conducted in this study reveal that major problems posed by uncertainties and risks for LSP companies during the COVID-19 pandemic are: decrease in cash flow, lack of continuity in production and distribution operations, thus increasing the risks, suspension of mergers and acquisitions in the industry, and demand fluctuations in sub-sectors disrupting the supply chain. Also, findings disclose that cause-effect structure of uncertainties and risks can contribute to LSPs' by helping them to develop and review the employee welfare strategy to avoid workflow disruption, reestablish distribution planning against demand and supply disruptions, reanalyze customers' and suppliers' financial situations, find new alternatives in the supply chain network, reevaluate contracts and insurance coverages for force majeure, digitize communication and business processes, and ensure flexibility in working conditions and durations.

Supply chains and LSPs operating in today's complex environment (Nilsson, 2006) are vulnerable to the associated uncertainties and risks. According to several studies (Christopher and Lee, 2004; Sanchez-Rodrigues et al., 2010b; Sreedevi and Saranga, 2017) , uncertainty causes supply chain risk. Contrarily, Calatayud et al. (2017) assert that risks boost uncertainty in a complex environment such as logistics. We merge these two opposite claims by concluding that uncertainties and risks are intertwined. The major contribution of our study is threefold. First, we identify the uncertainties and risks that LSPs encounter during the COVID-19 pandemic and demonstrate prominent ones. Second, we unveil the intertwined cause-effect structure of uncertainties and risks. Third, we provide an uncertainty and risk assessment guideline for LSPs operating in uncertain business environments shaped by the threats that emerged from unprecedented crises such as the COVID-19 pandemic.

Our study proposes a decision-making framework for LSPs by incorporating uncertainty and risk factors that affect their vulnerability in case of adverse events such as outbreaks, natural disasters, economic crises, regional conflicts, and force majeure. Our study combines the qualitative and MCDM methods to identify uncertainties and risks that LSPs encounter during the COVID-19 pandemic and investigate their prominence and cause-effect structure. At first, we conduct in-depth interviews and identify the uncertainties and risks via qualitative thematic analysis. Then, we collect data from LSP managers regarding the crippling effects of the novel coronavirus pandemic on the logistics industry and analyze their opinions on pairwise relations among uncertainties and risks by conducting the fuzzy DEMATEL method. We expose uncertainties and risks' prominence and interrelations and cluster them into net causer and net receiver groups. Afterward, we illustrate the cause-effect structure of the uncertainties and risks and provide an uncertainty and risk assessment tool for LSP companies. Our framework assists managers and decision-makers in allocating resources that require greater attention to companies in response to contingencies such as the COVID-19 pandemic. The proposed methodological framework in our study provides insights to the managers and decision-makers on how to prioritize uncertainties and risks to mitigate the negative impacts of adverse events such as the COVID-19 pandemic. The implications provided to the managers by applying the proposed framework guide them through the strategic decisions on allocating resources to counter unforeseeable threats. Companies prepare for unexpected situations and create a more flexible, agile, and resilient logistics service infrastructure by creating more effective resource planning. These capabilities improve customer service and satisfaction and give LSPs a competitive edge to outperform their competitors.

The COVID-19 pandemic affects the LSPs unequally, although the pandemic's real impact on the global supply chains is unknown (Twinn et al., 2020) . Accordingly, some LSPs serving the e-commerce market positively experience the COVID-19 shock by increased volume in their operations, while others negatively experience the crisis by delivery delays, congestions, and higher transportation rates (Pitel, 2020) . In future studies, sector-specific samples (e.g., food, medical, automotive, appliances, textile) can be studied.

Our study is conducted with data collected from the managers of the LSPs operating in Turkey. Further research is needed to illustrate employees' perspectives and compare managers and employees. More than half of our sample consists of large-scale companies operating in different businesses, e.g., diversified segments, sectors, and markets, which are more resilient to threats (Twinn et al., 2020) . Likewise, large-scale LSPs manage risks associated with environmental and social sustainability issues better than the smaller ones (Multaharju et al., 2017) . Overseas logistics experience more problems during the COVID-19 pandemic due to longer lead times and higher delivery costs (Chenneveau et al., 2020) . Therefore, further research can reflect the effects of company size or logistics mode. Our study focuses on how LSP managers evaluate uncertainties and risks, and future studies are needed to investigate how LSPs control and mitigate risks during adverse events. The logistics service providers during the COVID-19 pandemic: The prominence and the cause-effect structure of uncertainties and risks Highlights  Identifying uncertainties and risks of LSPs during the COVID-19 pandemic  Investigating the prominence and cause-effect structure of uncertainties and risks  Discovering intertwined schemes of uncertainties and risks of LSPs  Providing an uncertainty and risk assessment guideline for LSPs affected by threats

Analyzing barriers of Smart Energy City in Accra with two-step fuzzy DEMATEL

Application of structural equation modelling to analyse the impacts of logistics services on risk perception, agility and customer service level

The influencing factors of logistics integration and customer service performance for value creation of port logistics firms

Managing in-country transportation risks in humanitarian supply chains by logistics service providers: Insights from the 2015 Nepal earthquake

Employee welfare and job satisfaction in the Sri Lankan hotel industry

Understanding Coronanomics: The economic implications of the coronavirus (COVID-19) pandemic

Robust rationality and decisions under severe uncertainty

Analysis of external barriers to remanufacturing using grey-DEMATEL approach: An Indian perspective. Resources, Conservation and Recycling

Vulnerability of international freight flows to shipping network disruptions: A multiplex network perspective

Fuzzy DEMATEL method for developing supplier selection criteria

Coronavirus and technology supply chains: How to restart and rebuild. McKinsey & Company

Risk management and coordination in service supply chains: Information, logistics and outsourcing

The mean-variance approach for global supply chain risk analysis with air logistics in the blockchain technology era

Risk analysis in logistics systems: A research agenda during and after the COVID-19 pandemic

How relevant is volatility forecasting for financial risk management?

Mitigating supply chain risk through improved confidence

Forecasting product returns for remanufacturing operations

How can your industry respond at the speed of COVID-19's impact? Ernst & Young

COVID-19: Managing supply chain risk and disruption

Ensuring the integrity of transportation and logistics during the COVID-19 pandemic

A review of supply chain risk management: definition, theory, and research agenda

Who takes risks when and why? Determinants of risk taking

An introduction to qualitative research

Fused latent models for assessing product return propensity in online commerce

Working with qualitative data

Sustainable supply chains for supply chain sustainability: Impact of sustainability efforts on supply chain risk

Interrelationships of risks faced by third party logistics service providers: A DEMATEL based approach

Intuitionistic fuzzy based DEMATEL method for developing green practices and performances in a green supply chain. Expert Systems with Applications

Estimation under purposive sampling

Risk management processes in supplier networks

Assessment of healthcare waste treatment alternatives using an integrated decision support framework

Toward a supply chain risk identification and filtering framework using systems theory

Balancing in-house and outsourced logistics services: Effects on supply chain agility and firm performance

Coronavirus (COVID-19/SARS-CoV-2) and supply chain resilience: A research note

Impact of supply chain risk on agility performance: Mediating role of supply chain integration

Understanding uncertainty and risk in management

Evaluating critical factors influencing the reliability of emergency logistics systems using multiple-attribute decision making

Production, operating risk and market uncertainty: A valuation perspective on controlled policies. IIE Transactions (Institute of Industrial Engineers)

Integrated spare parts logistics and operations planning for maintenance service providers

Rational and participative task allocation between the nursing staff and the logistics support service provider in healthcare

The effect of logistics outsourcing on the supply chain vulnerability of shippers

Fuzzy method in group decision making

A value-created system of science (technology) park by using DEMATEL

A dominant maintenance strategy assessment model for localized third-party logistics service under Performance-Based consideration

A novel hybrid multiple criteria decision making model for material selection with target-based criteria

China's logistics development trends in the post COVID-19 era

A scheduling model of logistics service supply chain based on the mass customization service and uncertainty of FLSP's operation time

Quality control game model in logistics service supply chain based on different combinations of risk attitude

Risk identification map for a fashion retail supply chain. Proceedings of Summer School "Francesco Turco

COVID-19: Implications for business

A transformative supply chain response to COVID-19

Risks in emerging markets: Logistics services in the Mediterranean region

Sustainability-related risk management in buying logistics services

An operational risk analysis model for container shipping systems considering uncertainty quantification

Logistics management in practice -towards theories of complex logistics

A novel multiple criteria decision-making approach based on fuzzy DEMATEL, fuzzy ANP and fuzzy AHP for mapping collection and distribution centers in reverse logistics

Knightmare uncertainty: In the COVID-19 world, risk has become riskier

Defuzzification within a multicriteria decision model

Manufacturer's return policies and retail competition

Turkey's logistics providers adjust to the strains of COVID-19

Reason-code based model to forecast product returns

A critical analysis of supply chain risk management content: A structured literature review

Assessment of supply chain risk: Scale development and validation

Identifying causal relationships in strategy maps using ANP and DEMATEL

An integrated framework for the assessment of inbound supply risk and prioritization of the risk drivers: A real-life case on electronics supply chain

The role of physical distribution services as determinants of product returns in Internet retailing

Many (un)happy returns? The changing nature of retail product returns and future research directions

Risk and uncertainty in the insurance industry

Evaluating the causes of uncertainty in logistics operations

The impact of logistics uncertainty on sustainable transport operations

Establishing a transport operation focused uncertainty model for the supply chain

A combined big data analytics and Fuzzy DEMATEL technique to improve the responsiveness of automotive supply chains

Interplay between uncertainty and flexibility across the value-chain: Towards a transformation model of manufacturing flexibility

Application of fuzzy DEMATEL method for analyzing occupational risks on construction sites

A novel classification of supply chain risks: Scale development and validation

Managing uncertainty during a global pandemic: An international business perspective

Agriculture supply chain risks and COVID-19: mitigation strategies and implications for the practitioners

DEMATEL Technique: A Systematic Review of the State-of-the-Art Literature on Methodologies and Applications

Supply-chain uncertainty: a review and theoretical foundation for future research

Uncertainty and supply chain risk: The moderating role of supply chain flexibility in risk mitigation

Uncertainty and risk are multidimensional: Lessons from the COVID-19 pandemic

Logistics and cloud computing service providers' cooperation: A resilience perspective

Hybrid simulation-analytical modeling approaches for the reverse logistics network design of a third-party logistics provider

Prioritization of Facilitation Strategies of Park and Recreation Agencies Through DEMATEL Analysis

A hybrid approach for quantifying supply chain risk and prioritizing the risk drivers: A case of Indian petroleum supply chain

Real options: applications to logistics and transportation

Risk and uncertainty

The research methods knowledge base

Risk sensitivity in distribution channel partnerships: Implications for manufacturer return policies

Weighing risk and uncertainty

The impact of COVID-19 on logistics. International Finance Corporation

New concepts and trends of hybrid multiple criteria decision making

Fuzzy decision maps: A generalization of the DEMATEL methods

Evaluating intertwined effects in e-learning programs: A novel hybrid MCDM model based on factor analysis and DEMATEL

Research methods for everyday life: Blending qualitative and quantitative approaches

Mobile business intelligence: Key considerations for implementations projects

Joint logistics and financial services by a 3PL firm: Effects of risk preference and demand volatility

Logistics 4.0 solution-new challenges and opportunities

Impacts of supply chain uncertainty and risk on the logistics performance

Logistics innovation capability and its impacts on the supply chain risks in the Industry 4.0 era

The Measurement Model of Supply Chain Uncertainty and Risk in the Australian Courier Industry. Operations and Supply Chain Management

Evaluating logistics capability for mitigation of supply chain uncertainty and risk in the Australian courier firms

Uncertainty handling in containerized logistics: Unitary Traceability Object approach

Choosing knowledge management strategies by using a combined ANP and DEMATEL approach

Developing global managers' competencies using the fuzzy DEMATEL method

An integrated product returns model with logistics and marketing coordination

Flexibility and quality in logistics and relationships. Industrial Marketing Management

A combined fuzzy DEMATEL and TOPSIS approach for estimating participants in knowledge-intensive crowdsourcing

Sustainable recycling partner selection using fuzzy DEMATEL-AEW-FVIKOR: A case study in small-and-medium enterprises (SMEs)

Identifying critical success factors in emergency management using a fuzzy DEMATEL method

Writing -Review & Editing Sercan Demir: Conceptualization, Methodology, Validation, Formal Analysis, Writing -Original Draft Mehmet Akif Gunduz: Conceptualization, Methodology, Validation, Formal Analysis, Writing -Original Draft Fatih Cura: Resources, Writing -Original Draft