Lean, Agile, Resilient, Green, and Sustainable (LARGS) Supplier Selection Using Multi-Criteria Structural Equation Modeling under Fuzzy Environments

Abstract

This study aims to propose an integrated approach for supplier selection based on the lean, agile, resilience, green, and sustainable (LARGS) paradigm. This approach was validated using structural equation modelling (SEM) and the intuitionistic fuzzy TOPSIS method. A comprehensive literature review was conducted, identifying twenty-six criteria, which were then consolidated into five main criteria. A questionnaire was distributed to 237 individuals from manufacturing companies listed on the Tehran Stock Exchange, and the collected data were analyzed using third-order factor analysis and the partial least squares method. Subsequently, the proposed integrated approach was applied to evaluate four suppliers in an intuitionistic fuzzy environment, utilizing expert opinions and a case study on the automotive industry. The results demonstrate the effectiveness and practicality of the proposed approach in terms of prioritizing and selecting suitable suppliers according to LARGS criteria. In conclusion, this study contributes to the existing literature by proposing an integrated approach that addresses the decision-making challenges in supplier selection. This approach offers a practical tool for managers seeking to enhance sustainable supply chain performance from the LARGS perspective.

1. Introduction

In order to stay ahead in today’s competitive business landscape, it is crucial for companies to collaborate closely with external partners. Therefore, their attention should be directed towards the identification and selection of reliable suppliers. The selection of suppliers plays a major role in determining the competitiveness of an entire supply chain network [1]. The process of selecting suppliers is a vital part of strategic operations, and making incorrect decisions in this area can lead to severe and wide-ranging consequences [2]. Supplier management is considered a critical aspect of supply chain management, as it has implications for the overall performance and long-term relationships of a company. It is particularly significant for companies to allocate a significant portion of their sales revenue towards purchasing raw materials [3,4]. Hence, selecting the right suppliers is a crucial decision-making process aimed at minimizing purchasing risks, maximizing the overall value of purchases, and fostering enduring and close relationships with suppliers [5,6].

In the meantime, the availability of different suppliers with different capabilities complicates the selection process. Supplier selection is a process during which the best combination for meeting firm needs is selected from among the existing potential suppliers [7]. The process of selecting the best supplier, carried out to reduce the number of suppliers, can potentially lead to a competitive advantage for the manufacturer. This is achieved through attaining reduced costs, improved quality, and advancements in both process and product development. In fact, the main benefit of reducing the number of suppliers is that it provides more time to establish closer relationships with the remaining suppliers [8,9,10]. Scholars have employed various methods, such as Markov chains, deterministic and stochastic optimization, Bayesian networks, and simulation, for supplier selection. However, the topic of sustainable supplier selection in terms of resilience and sustainability has not been explored in previous studies [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99].

Baki [2] proposed a model using structural equation modeling and fuzzy ARAS to address the issue of green supplier selection, incorporating eight primary criteria and twenty-seven sub-criteria that considered classical, green, and social aspects. Sonar et al. [11] identified 12 key criteria influencing supplier selection within the LARGS paradigm and examined the interconnections between these criteria using interpretive structural modeling. Sahu et al. [100] proposed 63 measures for supplier selection based on LARGS (Lean, Agile, Resilient, Green, and Sustainable) criteria and utilized an integrated MCDM approach combining AHP, DEMATEL, ANP, Extended MOORA, and SAW techniques to assess suppliers. Tavana et al. [101] employed fuzzy group BWM and the FCoCoSo method for supplier selection in a reverse supply chain, focusing on LARG factors. Sheykhzadeh et al. [102] conducted a case study using fuzzy BWM and Fuzzy ARAS to evaluate the role of lean, agile, resilient, and green (LARG) criteria in the operations of a pharmaceutical supply chain distributor. Alimohammadlou and Khoshsepehr [103] developed hesitant fuzzy BWM, HFEDAS, HFCODAS, HFARAS, HFCOPRAS, HFWASPAS, and HFVIKOR based on green–resilient factors in the supplier selection process. Anvari [60] proposed a conceptual framework to investigate the integration of LARGS paradigms in the context of supply chain management and subsequently applied an integrated approach based on fuzzy AHP and fuzzy VIKOR to prioritize LARGS factors.

While the existing literature offers numerous analytical models for supplier selection, our research reveals a significant gap in the integration of structural equation modelling (SEM) and intuitionistic fuzzy Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) analysis with the use of lean, agile, green, resilient, and sustainable (LARGS) criteria. To address this gap, our study aims to develop an innovative approach that combines SEM and intuitionistic fuzzy TOPSIS analysis for sustainable supplier selection. SEM and intuitionistic fuzzy TOPSIS were selected because of their distinct advantages in addressing the complexities associated with LARGS criteria in supplier selection. SEM enables the assessment and estimation of causal relationships by utilizing both quantitative and qualitative data. The utilization of partial least squares (PLS)-based SEM provides a powerful and flexible approach for modeling complex data relationships, leading to reliable and accurate outcomes. PLS is particularly well suited for analyzing data exhibiting high dimensionality and multicollinearity and with small sample sizes. Therefore, incorporating PLS within the SEM framework enhances the rigor and accuracy of our approach, ensuring that the findings derived from our study can be trusted and applied in practice. On the other hand, intuitionistic fuzzy TOPSIS is a multi-criterion decision-making (MCDM) approach that effectively handles decision making under conditions of uncertainty and imprecision, making it suitable for tackling the inherent uncertainties in sustainable supplier selection processes. By addressing the research gap related to the integration of SEM, intuitionistic fuzzy TOPSIS, and LARGS criteria, our study contributes to the advancement of knowledge on sustainable supplier selection. This integration is crucial, as it enhances the current understanding of how organizations can effectively evaluate and select suppliers based on LARGS criteria, ultimately promoting sustainable practices throughout the supply chain. In addition, many scholars have attempted to use MCDM methods combined with classical fuzzy concepts to overcome uncertainties and ambiguities in decision-making problems [12,13,14]. However, since the development of fuzzy methods such as intuitionistic fuzzy sets is an efficient, effective, and flexible method for investigating decision-making problems, the present study intends to use TOPSIS combined with an intuitionistic fuzzy set approach to solve the supplier selection problem by integrating lean, agile, green, resilient, and sustainable aspects of the supply chain. The validity if this approach was confirmed by applying the SEM method in a case study.

The remaining parts of this paper are structured as follows: Section 2 reviews the relevant literature. Section 3 describes the proposed methodology. In Section 4, the results obtained using the proposed methodology are discussed and tested using a sensitivity analysis. Finally, Section 5 concludes our work.

2. Literature Review

2.1. Lean Supply Chain

Leanness involves the creation of a value stream with the aim of eliminating waste, addressing aspects including time, inventory, unnecessary costs, and improving production planning [15]. The primary focus of lean thinking is to remove any waste or excess materials for an organization. According to the lean system perspective, anything that exceeds the minimum requirement and does not add value to the product, such as production factors, materials, human resources, parts, machinery, and time, is considered waste and should be eliminated [16]. The ultimate objective of lean production is to achieve better results with fewer labor hours and lower costs. This approach is most effective in environments with relatively stable and predictable demand and limited product variety [17,18,19,104].

In the lean supply chain model, enhancing the effectiveness of the production system and minimizing machine setup time are crucial elements. These factors contribute to economic production in smaller batches, decreasing prices, improving profitability, and greater production flexibility [20]. Although the internal flexibility of a chain increases with the reduction in setup time in the lean supply chain, flexibility and responsiveness to customer demand in product design, scheduling, and distribution are also of great importance and yet scarcely considered in lean production [21].

2.2. Agile Supply Chain

The lean paradigm is known for its application in maximizing profits while reducing costs, while the agile paradigm tends to involve maximizing profits by responding quickly to customer demands [11,17,22]. In order to survive in dynamic and changing markets, supply chains need a tool that can overcome environmental challenges. This tool is agility. The main success factors in an agile supply chain include the application of information technology, the integration of processes, appropriate planning, the development of employee skills, sensitivity and responsiveness to the market, new product launches, flexibility, delivery speed, cost reduction, product quality, and customer satisfaction [23,24].

The concept of an agile supply chain emphasizes the importance of adaptability and flexibility in responding quickly and effectively to changing markets [25]. To be competitive, an industrial organization’s supply chain must have both capable and competitive components as well as incorporate agility. This means having competitive suppliers with global capabilities and the ability to adapt to changes. An agile supply chain is focused on managing change, uncertainty, and unpredictability in its business environment and providing appropriate responses to these challenges. To achieve this, specific capabilities or competencies are required [26]. These capabilities can be grouped into four categories: (1) responsiveness, which involves recognizing and quickly responding to changes, both reactively and proactively, and leveraging changes to improve and expand; (2) competence, which involves effectively and efficiently achieving organizational goals; (3) flexibility/adaptability, which involves utilizing different processes and facilities to achieve similar goals; and (4) speed, which involves completing activities as quickly as possible [27,28].

2.3. Resilient Supply Chain

Supply chain resilience refers to a supply chain’s adaptability in regard to preparing for unexpected events, responding to disturbances, maintaining continuity of operations at an optimal level, and controlling structure and performance [28,29]. Resilient supply chains are capable of delaying disruptions and mitigating their impact, resisting acceptable destruction parameters, and effectively recovering within an acceptable timeframe and with respect to cost and risk [30,31,32]. However, supply chains have more than just a reactive ability; they also enable supply chain members to withstand problems and adversities and seize opportunities in varied situations [33]. Consequently, supply chain resilience provides companies with a competitive advantage over their rivals by allowing them to position themselves better during disruptions [28,31]. According to Carvalho and Cruz-Machado [17,22], resilience is a category that needs to be developed. They argue that cooperation, flexibility, and transparency should be established among different groups of suppliers and customers. Resilient supply chains are capable of effectively adapting to disturbances while maintaining the same level of efficiency [34].

2.4. Green Supply Chain

The concept of green supply chain management emerged during the early 1990s, with its origins tracing back to Michigan State University. As environmental management gained traction, there was a rise in studies focused on biocompatible production strategies and green supply chain management. Over the past decade, the emergence of green supply chains has had a significant impact, allowing companies to align their supply chains with environmental objectives [10,35,36]. A green supply chain refers to the various measures, taken both within and outside a firm’s supply chain, aimed at preventing pollution and promoting environmental improvements. It presents a valuable opportunity for those interested in sustainable consumption and environmental business practices [37]. To ensure long-term success, companies need to reassess their product design and production methods in order to develop environmentally friendly products. It is crucial for companies to effectively manage and coordinate their relationships with suppliers, specifically their respective green supply chains [38,39].

Green supply chain management activities are a tool for environmental protection best viewed as consisting of inter-organizational activities that allow supply chain members to cooperate throughout a chain and preserve the environment [40,41]. The goal of implementing green supply chain management in business activities is to simultaneously improve economic and environmental performance. Economic improvement is reflected by reducing the cost of material purchasing, reducing energy consumption, and reducing waste emissions. Besides the direct and explicit financial benefits, such as cost savings and increased profits, manufacturers can also obtain other advantages from green activities, such as readiness to produce new products and adhere to new regulations, improved relations with customers, a green corporate image, and better social responsibility [37,42]. In fact, the green supply chain concept focuses on the incorporation of environmental management and supply chain management in order to mitigate the environmental impacts throughout the product life cycle through information sharing and the collaborative efforts of all members involved in the supply chain [43].

2.5. Sustainable Supply Chain

The concept of a sustainable supply chain includes managing the flow of capital, information, and materials while promoting collaboration among companies within the supply chain. This approach integrates objectives from the economic, social, and environmental dimensions of sustainable development, which are driven by the needs of customers and stakeholders [10,44]. To ensure these objectives are being implemented throughout the supply chain, members of sustainable supply chains adopt social and environmental criteria. At the same time, they strive to maintain competitiveness by meeting customer demands and economic criteria [45,46]. Therefore, environmental, social, and cultural concerns have prompted scholars and scientists to conduct research in order to develop and present new approaches in supply chains that lead to sustainable development [47]. In terms of sustainability, the three economic, environmental, and social aspects are fully interconnected, and most scholars examine sustainability from these three perspectives [48]. According to the authors of [49], sustainable supply chain management involves both internal actions like developing sustainable products and processes and external actions such as involving suppliers and customers to enhance a supply chain’s sustainability.

In a sustainable chain, the environmental goal is part of the supplier selection criteria. Incorporating these environmental criteria into the evaluation of suppliers allows only a very limited number of suppliers to meet the defined criteria [11]. The concept of sustainable supply chains is widely acknowledged as an integral aspect of promoting industrial sustainability, aiming to minimize ecological footprints, lower expenses, satisfy customer demands, and fulfill relevant economic standards [11,50,51,52].

In summary, previous research has highlighted the significance of lean, agile, resilient, green, and sustainable (LARGS) supply chains in achieving success in supply chain management. However, a review of the existing studies reveals a lack of models or frameworks addressing the roles of these factors specifically in supplier selection. Consequently, the primary objective of this study was to develop a model that examines the impact of the LARGS paradigm on supplier selection. The criteria for evaluating suppliers based on the LARGS paradigm, as found in the literature, are outlined in

Table 1. Supplier selection criteria based on the literature.

Criterion	Sub-Criterion	References
Lean	Price	[10,11,53,54,55,56]
	Lead time	[11,54,57,58,59]
	Defect reduction	[10,56,60,61]
	Just-in-time practices	[62,63,64]
	Collaboration with SC	[11,65]
Agile	Innovation capability	[11,66,67,68,69]
	Response speed	[11,62,70]
	Dynamic alliance	[62,71,72]
	New production line capability	[11,73,74]
	Flexibility	[11,54,68,70]
Resilient	Ability to deal with unexpected disruptions	[15,29,75]
	Resilient transportation	[22,29,62,72]
	Strategic capacity and inventory buffer	[11,15,76]
	SC risk management culture	[77,78]
	Capacity for mass customization	[11,62,79]
Green	Green Innovation	[60,69,80,81]
	Green product	[1,2,82]
	Environment management system	[10,53,54,83]
	Renewable energy/initiative	[62,63,71]
	Waste recycling	[62,72,84]
	Environmental competency	[2,85,86]
Sustainable	Life cycle assessment system	[79,80,87,88]
	Long-term relationships	[11,83,89,90]
	Sustainabfle measuring criteria	[10,84,91]
	Reputation	[11,83,87,89]
	Cooperation among various entities in a SC	[60,91]

.

3. Research Methodology

The methodology used in the present study involves third-order factor analysis and partial least squares for analyzing the proposed conceptual model. Third-order factor analysis is a statistical method that aids the identification of latent variables, which are constructs that cannot be directly measured but can be inferred based on observed variables. This method was chosen because it allowed us to model complex relationships between observed variables (26 criteria for supplier selection) and latent factors, thereby providing a comprehensive insight into the underlying structure of our data. Moreover, the third-order factor analysis helped us to identify the underlying latent variables, while PLS enabled us to model the structural relationships between these factors and the observed variables. Also, for supplier selection and evaluation, intuitionistic fuzzy TOPSIS analysis was applied. Additionally, the weights assigned to the experts for evaluating suppliers using the Intuitionistic Fuzzy TOPSIS method were not equal.

3.1. Population and Sample

To ensure the representativeness of the sample, specific criteria were employed. The statistical population comprised managers and deputies of the supply chains in large and medium-sized manufacturing companies listed on the Tehran Stock Exchange, including those from the automobile, car-utility, pipe-and-faucet, electric utility, clothing, food, chemical, mining, pharmaceutical, and petrochemical industries. A random selection process was used to select 200 companies with more than 50 employees from the entire pool of manufacturing companies. This approach was designed to minimize bias and ensure transparency, thereby increasing the generalizability of the findings to the larger population. Efforts were made to contact the supply chain managers and deputies of the selected companies based on their roles and responsibilities within the supply chain. This ensured that the respondents possessed the knowledge and expertise necessary for providing insightful responses. However, potential limitations, such as response bias or the exclusion of individuals who were not available or willing to participate, should be acknowledged. To achieve a representative sample, 330 questionnaires were distributed among these supply chain managers and deputies of the selected companies. Of these, 237 (72%) were filled out and returned, providing valuable insights into the research objectives. In the second phase, a case study on the automotive industry was conducted to evaluate and select suppliers.

3.2. Data Collection Instruments

A questionnaire with 87 items was developed to measure the identified variables in the first phase. A five-point Likert scale, ranging from completely disagree (1) to completely agree (5), was used to measure all items to attain reliable and valid responses. The five-point Likert scale is widely used in social science research and is suitable for measuring attitudes, opinions, and perceptions. Two separate questionnaires were utilized in the subsequent phase, focusing on the importance of criteria and the ratings of suppliers, respectively. The use of a five-point Likert scale ensured consistent measurement across items, enhancing the reliability and validity of the collected data. This approach facilitated a focused data collection process, contributing to the rigor of this study.

4. Results

Our structural model utilizes a third-order factor analysis with three hierarchical levels. At the first-order level, a combination of indicators or questions accounts for lean components (e.g., price, lead time, just-in-time practices, defect reduction, and collaboration with the supply chain), agile components (e.g., innovation capability, response speed, dynamic alliance, new production line capability, and flexibility), resilient components (e.g., the ability to deal with unexpected disruptions, resilient transportation, strategic capacity and inventory buffer, supply chain risk management culture, and capacity for mass customization), green components (e.g., green innovation, green products, environment management systems, renewable energy/initiatives, waste recycling, and environmental competence), and sustainable components (e.g., life cycle assessment systems, long-term relationships, sustainable measuring criteria, reputation, and cooperation among various entities in the supply chain). The amalgamation of these indicators yielded five components—lean, agile, resilient, green, and sustainable—comprising the second order of the model. Finally, from the combination of these five components, a general variable named LARGS supplier selection was obtained.

To delve into the validation process, we assessed construct validity by examining the factor loadings in the confirmatory factor analysis. A factor loading exceeding 0.6 for each indicator with its construct indicates that each item adequately represents said construct [92,93]. The factor loadings are presented in

Table 2. Factor loadings, CR, and AVE of variables.

	Variable	Item	Factor Loading	Cronbach’s Alpha	CR	AVE
Lean	Price	1	0.787	0.724	0.845	0.656
		2	0.815
		3	0.807
	Lead time	1	0.793	0.685	0.826	0.613
		2	0.773
		3	0.782
	Defect reduction	1	0.792	0.703	0.835	0.628
		2	0.834
		3	0.750
	Just-in-time practices	1	0.837	0.799	0.869	0.624
		2	0.767
		3	0.814
		4	0.749
	Collaboration with SC	1	0.698	0.734	0.834	0.557
		2	0.781
		3	0.777
		4	0.726
Agile	Innovation capability	1	0.831	0.792	0.878	0.706
		2	0.801
		3	0.823
		4	0.824
	Response speed	1	0.803	0.837	0.891	0.672
		2	0.860
		3	0.856
	Dynamic alliance	1	0.865	0.787	0.876	0.703
		2	0.783
		3	0.864
	New production line capability	1	0.760	0.827	0.896	0.744
		2	0.905
		3	0.914
	Flexibility	1	0.888	0.831	0.899	0.748
		2	0.875
		3	0.830
Resilient	Ability to deal with unexpected disruptions	1	0.834	0.804	0.884	0.718
		2	0.856
		3	0.852
	Resilient transportation	1	0.901	0.899	0.937	0.833
		2	0.875
		3	0.960
	Strategic capacity and inventory buffer	1	0.766	0.827	0.885	0.658
		3	0.817
		3	0.850
		4	0.809
	SC risk management culture	1	0.833	0.80	0.870	0.626
		2	0.743
		3	0.788
		4	0.799
	Capacity for mass customization	1	0.736	0.74	0.809	0.585
		2	0.786
		3	0.773
Green	Green Innovation	1	0.778	0.734	0.850	0.653
		2	0.821
		3	0.825
	Green products	1	0.834	0.838	0.892	0.674
		2	0.837
		3	0.835
		4	0.774
	Environment management systems	1	0.858	0.856	0.902	0.698
		2	0.786
		3	0.837
		4	0.860
	Renewable energy/initiatives	1	0.852	0.848	0.908	0.766
		2	0.896
		3	0.878
	Waste recycling	1	0.891	0.814	0.890	0.730
		2	0.840
		3	0.831
	Environmental competency	1	0.70	0.752	0.845	0.586
		2	0.816
		3	0.852
		4	0.840
Sustainable	Life cycle assessment systems	1	0.874	0.805	0.886	0.722
		2	0.894
		3	0.777
	Long-term relationships	1	0.834	0.726	0.850	0.656
		2	0.864
		3	0.725
	Sustainable measuring criteria	1	0.743	0.832	0.888	0.666
		2	0.798
		3	0.872
		4	0.845
	Reputation	1	0.865	0.866	0.918	0.789
		2	0.918
		3	0.881
	Cooperation among the various entities in the SC	1	0.880	0.814	0.890	0.729
		2	0.831
		3	0.850

. In the structural equation model, both construct and discriminant validity are crucial. Construct validity assesses the importance of the chosen indicators in measuring the constructs, while discriminant validity ensures that the indicators of each construct effectively differentiate themselves from those of other constructs. The average variance extracted (AVE) is utilized to assess discriminant validity, with Fornell and Larcker [94] suggesting that a threshold of 0.50 or higher should be set. As shown in Table 2, all the studied constructs have AVE values exceeding 0.5.

Reliability was evaluated through composite reliability and Cronbach’s alpha coefficient. Values above 0.7 for each construct demonstrate satisfactory reliability [93]. Table 2 displays the factor loadings, composite reliability, and AVE of the variables, indicating robust validity and reliability across the constructs.

Utilizing SmartPLS4 software for its ability to analyze second-order factor models, we explored latent factors influenced by a higher-level latent variable. The second-order factor analysis results, including the factor loadings and t-values for indicators linked to each construct, are presented in

Table 3. Factor loadings and t-values for determining significance and validating LARGS indicators.

Variable		Factor Loading	t-Value	Result
Lean	Price	0.798	32.201	Confirmed
	Lead time	0.840	35.404	Confirmed
	Defect reduction	0.747	23.093	Confirmed
	Just-in-time practice	0.845	38.863	Confirmed
	Collaboration with SC	0.848	40.377	Confirmed
Agile	Innovation ability	0.878	47.944	Confirmed
	Response speed	0.851	40.692	Confirmed
	Dynamic alliance	0.839	40.998	Confirmed
	Capacity for new product lines	0.781	26.167	Confirmed
	Flexibility	0.832	41.731	Confirmed
Resilient	Capacity for dealing with unexpected disruptions	0.799	25.882	Confirmed
	Flexible transportation	0.765	24.944	Confirmed
	Strategic capacity and inventory buffer	0.823	11.936	Confirmed
	SC risk management culture	0.889	65.366	Confirmed
	Capacity for mass customization	0.836	35.352	Confirmed
Green	Green innovation	0.848	37.795	Confirmed
	Green product	0.875	46.481	Confirmed
	Environment management systems	0.857	38.267	Confirmed
	Renewable energy/initiatives	0.746	29.273	Confirmed
	Waste recycling	0.862	40.405	Confirmed
	Environmental competency	0.879	44.835	Confirmed
Sustainable	Lifecycle assessment systems	0.848	31.316	Confirmed
	Long-term relationships	0.837	28.171	Confirmed
	Sustainable measuring criteria	0.898	55.541	Confirmed
	Reputation	0.858	46.708	Confirmed
	Cooperation among the various entities in SC	0.849	45.126	Confirmed

. All the components of LARGS exhibit factor loadings exceeding 0.7 and t-values exceeding 1.96 (p > 0.05), confirming the significance and relevance of the chosen indicators.

Transitioning to a third-order factor analysis, the lean, agile, resilient, green, and sustainable components each have indicators serving as markers, necessitating this analysis. The questions were subjected to an analysis using third-order factor analysis, and the findings are provided below. For path coefficients and R² values, the Partial Least Squares (PLS) method was employed to examine the theoretical model and hypotheses. T-values were calculated using the bootstrap method, with 500 sub-samples. Factor loadings, R² values, and Stone-Geisser’s Q2 [95] coefficient are presented in

Table 4. Third-order CFA results for supplier selection.

Path	Factor Loading	t-Value	p	R2	Q2
Effect of leanness on supplier selection	0.875	44.418	0.001	0.766	0.307
Effect of agility on supplier selection	0.894	65.419	0.001	0.798	0.391
Effect of resilience on supplier selection	0.904	59.122	0.001	0.817	0.371
Effect of greenness on supplier selection	0.938	107.289	0.001	0.879	0.435
Effect of sustainability on supplier selection	0.906	62.257	0.001	0.82	0.423

, indicating significant explained variances for the lean (0.766), agile (0.798), resilient (0.817), green (0.879), and sustainable (0.820) components. Positive Q2 values suggest substantial predictability. Table 4 reveals positive and significant effects of leanness (β = 0.875, p < 0.001), agility (β = 0.894, p < 0.001), resilience (β = 0.904, p < 0.001), greenness (β = 0.938, p < 0.001), and sustainability (β = 0.906, p < 0.001) on supplier selection. Figure 1 illustrates the tested model, depicting the factors influencing supplier selection, with the numbers within the circles representing the magnitude of explained variances.

In general, the goodness-of-fit (GOF) index in PLS serves as a measure for assessing the validity and quality of a PLS model. It evaluates the model’s collective predictive capacity and determines if it successfully predicts the endogenous variables. In this study, the tested model displayed a GOF value of 0.67, indicating a good fit. Values exceeding 0.36 suggest satisfactory and acceptable levels of quality for a model [96].

5. Evaluation of Suppliers in a Case Study

This research paper focuses on a case study of the supply chain department in an Iranian automotive company. This department is responsible for providing parts to the production lines of an automotive manufacturing company. Its primary objective is to identify and select automotive parts suppliers for both domestic and international markets. The main challenge faced by this department concerns choosing the most suitable supplier and allocating demand suppliers under uncertain environments, using LARGS indicators that align with the company’s requirements. To evaluate suppliers, a committee comprising four experts with extensive experience in the automotive industry (each having over 10 years of work experience) was formed. The committee consisted of a supply chain manager, a procurement manager, a specialized purchasing expert, and a specialized logistics expert. As shown in Figure 2, four suppliers were considered in the study, taking into account 26 criteria that had significant and positive effects, as confirmed using the SEM method. Four suppliers were considered in the study, taking into account 26 criteria that had significant and positive effects, as confirmed using the SEM method. In this study, it is supposed that there is no dependency between the criteria. Subsequently, the following section presents the results of the proposed steps for implementing the intuitionistic fuzzy TOPSIS analysis introduced by Rouyendegh [97] in a case study.

• Step 1. Calculation of the relative importance of Decision Makers (DMs).

Initially, the level of significance for each decision maker was established by utilizing a designated linguistic variable and employing intuitionistic fuzzy numbers (IFNs) in accordance with the information presented in

Table 5. Linguistic terms for the importance of decision makers and criteria.

Linguistic Terms	IFNs
Very important (VI)	0.90	0.10	0.00
Important (I)	0.75	0.20	0.05
Medium (M)	0.50	0.45	0.05
Unimportant (U)	0.35	0.60	0.05
Very unimportant (VU)	0.10	0.90	0.00

.

Assuming that the importance of each decision maker is based on IFNs represented as Dk = (μ_k,v_k,π_k), where is the degree of membership function, is the degree of non-membership function, and is the hesitation degree, then the final weight of the k-th DM can be determined as shown in (1).

$${\lambda }_k=\frac{\left({\mu }_k+{\pi }_k(\frac{{\mu }_k}{{\mu }_k+v_k})\right)}{{\displaystyle \sum _{k=1}^K\left({\mu }_k+{\pi }_k(\frac{{\mu }_k}{{\mu }_k+v_k})\right)}};{\displaystyle \sum _{k=1}^K{\lambda }_k=1}$$

(1)

DM1 and DM2 were classified as “very important,” DM3 was classified as “important,” and DM4 was classified as being of “medium” importance. Subsequently, the determination of the importance weights for these decision makers was carried out. The importance weights of DM1 to DM4 are 0.289, 0.289, 0.253, and 0.169, respectively.

• Step 2. Determination of criteria weights based on decision makers’ judgments

The group decisions were aggregated using Equation (2) by applying the Intuitionistic Fuzzy Weighted Averaging (IFWA) operator proposed by Xu [98] based on the linguistic terms and intuitionistic fuzzy numbers presented in Table 5.

$$\begin{array}{l}{W}_j=IFW{A}_{\lambda }(w_j^{(1)},w_j^{(2)},\dots ,w_j^{(k)})={\lambda }_1{w}_j^{(1)}\oplus {\lambda }_1{w}_j^{(2)}\oplus ,\dots ,\oplus {\lambda }_1{w}_j^{(k)}\\ =[1-\underset{k=1}{\overset{K}{{\displaystyle \Pi }}}{(1-{\mu }_{ij}^{(k)})}^{{\lambda }_k},\underset{l=1}{\overset{k}{{\displaystyle \Pi }}}(v_{ij}^{(k)}),\underset{l=1}{\overset{K}{{\displaystyle \Pi }}}{(1-{\mu }_{ij}^{(k)})}^{{\lambda }_k}-\underset{l=1}{\overset{K}{{\displaystyle \Pi }}}{(1-v_{ij}^{(k)})}^{{\lambda }_k}\end{array}$$

(2)

The criteria’s importance is indicated by the linguistic terms listed in

Table 6. Weights of criteria.

Criteria	DM1	DM2	DM3	DM4	Calculated Weights
Price, C1	VI	I	VI	I	0.848	0.137	0.015
Lead time, C2	M	I	I	M	0.657	0.290	0.053
Defect reduction, C3	I	M	I	I	0.695	0.253	0.053
Just-in-time practices, C4	M	U	M	M	0.461	0.489	0.050
Collaboration with SC, C5	M	U	VU	U	0.346	0.612	0.042
Innovation ability, C6	I	I	M	I	0.702	0.246	0.052
Response speed, C7	I	VI	M	I	0.771	0.201	0.028
Dynamic alliance, C8	M	M	U	M	0.466	0.484	0.050
Capacity for new product lines, C9	M	U	M	VU	0.404	0.550	0.046
Flexibility, C10	I	VI	VI	VI	0.870	0.122	0.008
Capacity for dealing withunexpected disruptions, C11	VI	I	I	I	0.808	0.164	0.028
Flexible Transportation, C12	I	I	I	M	0.719	0.229	0.052
Strategic capacity and inventory buffers, C13	M	M	U	M	0.466	0.484	0.050
SC risk management culture, C14	I	I	VI	M	0.777	0.192	0.030
Capacity for mass customization, C15	U	M	VU	U	0.346	0.612	0.042
Green innovation, C16	M	I	M	M	0.591	0.356	0.053
Green products, C17	M	I	I	I	0.695	0.253	0.053
Environment management systems, C18 C18	I	I	I	M	0.719	0.229	0.052
Renewable energy/initiative, C19	M	M	M	U	0.477	0.472	0.050
Waste recycling, C20	VI	I	I	VI	0.836	0.146	0.019
Environmental competency, C21	M	M	U	I	0.525	0.422	0.053
Lifecycle assessment systems, C22	M	I	I	I	0.695	0.253	0.053
Long-term relationships, C23	VI	VI	I	I	0.853	0.134	0.013
Sustainable measuring criteria, C24	M	M	U	M	0.466	0.484	0.050
Reputation, C25	I	I	VI	I	0.802	0.168	0.030
Cooperation among the various entities in a SC, C26	U	M	VU	U	0.346	0.612	0.042

. Numerical representations of the linguistic terms can be found in Table 5 corresponding to step 2. The criteria weights were ultimately aggregated and are displayed in Table 6.

• Step 3. Creating the aggregated intuitionistic fuzzy decision matrix.

By referring to the numerical equivalents of the linguistic terms as provided in

Table 7. Linguistic terms for rating the alternatives.

Linguistic Terms	IFNs
Extremely high (EH)	1	0	0
Very-very high (VVH)	0.90	0.10	0.00
Very high (VH)	0.80	0.10	0.10
High (H)	0.70	0.20	0.10
Medium-high (MH)	0.60	0.30	0.10
Medium (M)	0.50	0.40	0.10
Medium-low (ML)	0.40	0.50	0.10
Low (L)	0.25	0.60	0.15
Very low (VL)	0.10	0.75	0.15
Very-very low (VVL)	0.10	0.90	0.00

, the ratings for the alternatives were determined. In order to achieve this, it was necessary to combine the individual opinions obtained from a group of DMs into a unified viewpoint. This unified viewpoint was then used to create an aggregated intuitionistic fuzzy decision matrix (IFDM). The IFWA method, as shown in

Table 8. Ratings of alternatives based on criteria and relevant aggregated IFNs.

Criteria	Suppliers	DM1	DM2	DM3	DM4	The Aggregated IFNs			Criteria	DM1	DM2	DM3	DM4	The Aggregated IFNs
C1	S1	EH	H	MH	H	1.000	0.000	0.000	C14	H	VH	H	MH	0.720	0.175	0.105
	S2	H	H	VH	MH	0.716	0.180	0.105		M	ML	H	MH	0.554	0.341	0.105
	S3	VH	H	MH	MH	0.699	0.194	0.107		L	ML	M	ML	0.389	0.498	0.113
	S4	VH	VVH	EH	VH	1.000	0.000	0.000		H	VVH	H	H	0.782	0.164	0.055
C2	S1	M	MH	ML	M	0.509	0.390	0.101	C15	L	L	ML	MH	0.363	0.510	0.128
	S2	L	M	M	MH	0.459	0.428	0.113		VL	VL	ML	L	0.213	0.652	0.136
	S3	MH	H	M	H	0.629	0.268	0.103		ML	ML	ML	M	0.418	0.482	0.100
	S4	VH	VH	H	MH	0.751	0.144	0.106		H	ML	M	M	0.545	0.349	0.106
C3	S1	ML	L	H	ML	0.463	0.418	0.119	C16	MH	M	M	ML	0.517	0.382	0.101
	S2	L	VL	L	L	0.209	0.640	0.151		L	M	M	L	0.398	0.482	0.120
	S3	M	M	ML	M	0.476	0.423	0.100		MH	ML	ML	H	0.525	0.370	0.105
	S4	L	VVL	VL	L	0.172	0.714	0.114		VH	EH	H	MH	1.000	0.000	0.000
C4	S1	VVH	VVH	H	VVH	0.868	0.119	0.013	C17	VH	H	H	H	0.733	0.164	0.103
	S2	EH	VVH	EH	H	1.000	0.000	0.000		MH	H	MH	M	0.618	0.280	0.102
	S3	H	M	MH	H	0.626	0.271	0.103		ML	ML	L	ML	0.365	0.524	0.111
	S4	M	MH	H	M	0.588	0.309	0.103		H	VH	VH	MH	0.747	0.147	0.106
C5	S1	EH	VH	VVH	EH	1.000	0.000	0.000	C18	H	VH	MH	H	0.713	0.181	0.106
	S2	H	H	VH	MH	0.716	0.180	0.105		ML	MH	MH	M	0.533	0.365	0.102
	S3	MH	MH	VH	VH	0.702	0.189	0.110		MH	ML	M	M	0.506	0.393	0.102
	S4	VVH	H	VVH	VH	0.846	0.122	0.032		VVH	H	H	MH	0.771	0.175	0.054
C6	S1	M	H	H	MH	0.635	0.262	0.103	C19	VH	H	H	VH	0.751	0.146	0.104
	S2	VH	VVH	H	VVH	0.839	0.119	0.042		VH	MH	MH	H	0.688	0.204	0.108
	S3	L	ML	L	M	0.343	0.532	0.125		M	ML	L	ML	0.398	0.491	0.111
	S4	ML	L	M	VVL	0.346	0.550	0.104		VVH	EH	EH	VVH	1.000	0.000	0.000
C7	S1	VH	VH	H	H	0.763	0.134	0.103	C20	L	VL	L	ML	0.239	0.621	0.141
	S2	M	MH	H	M	0.588	0.309	0.103		ML	M	M	ML	0.456	0.443	0.101
	S3	H	VH	H	H	0.733	0.164	0.103		MH	MH	M	H	0.597	0.301	0.102
	S4	H	H	VH	H	0.729	0.168	0.103		MH	H	MH	MH	0.632	0.267	0.101
C8	S1	M	MH	L	VL	0.426	0.454	0.120	C21	H	VH	MH	MH	0.699	0.194	0.107
	S2	VH	H	VH	MH	0.747	0.147	0.106		H	ML	MH	M	0.570	0.325	0.105
	S3	M	VH	VH	MH	0.707	0.180	0.113		L	ML	L	M	0.343	0.532	0.125
	S4	VVH	H	VVH	VH	0.846	0.122	0.032		VH	VH	H	VH	0.778	0.119	0.102
C9	S1	VH	H	MH	MH	0.699	0.194	0.107	C22	VH	H	MH	H	0.713	0.181	0.106
	S2	EH	VVH	H	VH	1.000	0.000	0.000		MH	H	H	MH	0.658	0.241	0.101
	S3	MH	ML	ML	ML	0.466	0.431	0.102		M	MH	MH	M	0.557	0.342	0.101
	S4	M	MH	M	VVL	0.482	0.422	0.096		VVH	EH	EH	VH	1.000	0.000	0.000
C10	S1	H	ML	M	ML	0.531	0.363	0.106	C23	EH	H	MH	MH	1.000	0.000	0.000
	S2	VL	VL	ML	L	0.213	0.652	0.136		M	L	MH	M	0.469	0.418	0.113
	S3	ML	M	ML	M	0.448	0.451	0.101		M	ML	M	L	0.436	0.457	0.108
	S4	M	ML	VL	ML	0.369	0.520	0.111		EH	VH	H	VH	1.000	0.000	0.000
C11	S1	EH	VVH	H	H	1.000	0.000	0.000	C24	MH	MH	H	MH	0.628	0.271	0.101
	S2	ML	ML	L	L	0.341	0.540	0.119		H	M	H	M	0.621	0.275	0.104
	S3	H	MH	M	H	0.629	0.268	0.103		ML	ML	M	M	0.444	0.455	0.101
	S4	H	EH	VH	VH	1.000	0.000	0.000		VVH	VH	EH	VH	1.000	0.000	0.000
C12	S1	M	ML	M	L	0.436	0.457	0.108	C25	VH	MH	MH	H	0.688	0.204	0.108
	S2	VH	H	H	VH	0.751	0.146	0.104		ML	M	MH	MH	0.520	0.378	0.102
	S3	L	VL	ML	L	0.253	0.611	0.136		M	ML	M	M	0.473	0.427	0.100
	S4	M	M	MH	MH	0.545	0.354	0.101		VVH	VH	H	EH	1.000	0.000	0.000
C13	S1	M	M	MH	MH	0.545	0.354	0.101	C26	EH	VH	VVH	EH	1.000	0.000	0.000
	S2	H	VH	VH	MH	0.747	0.147	0.106		H	H	VH	MH	0.716	0.180	0.105
	S3	L	VL	ML	M	0.302	0.571	0.127		MH	MH	VH	VH	0.702	0.189	0.110
	S4	H	H	VVH	VH	0.788	0.149	0.063		VVH	H	VVH	VH	0.846	0.122	0.032

, was utilized to calculate the IFDM, taking into account the preferences of the decision makers.

• Step 4. Calculating the aggregated weighted intuitionistic fuzzy decision matrix.

The computation of an aggregated weighted intuitionistic fuzzy decision matrix can be carried out by utilizing the definitions provided in Equation (3) after the determination of criteria weights (W) and the creation of an aggregated IFDM (R matrix), as stated by Memari [99].

$$R\otimes W=({{\mu }^{\prime }}_{ij},{v^{\prime }}_{ij})=\left\{\langle x,{\mu }_{ij}\times {\mu }_j,v_{ij}+v_j-v_{ij}\ast v_j\rangle \right\}$$

(3)

• Step 5. Deriving an intuitionistic fuzzy positive-ideal solution and a negative-ideal solution.

To calculate the intuitionistic fuzzy positive-ideal solution and intuitionistic fuzzy negative-ideal solution, the positive and negative criteria were denoted as J⁺ and J⁻, respectively.

$$A^{+}=({{\mu }^{\prime }}_j^{*},{{\nu }^{\prime }}_j^{*},{{\pi }^{\prime }}_j^{*}),A^{-}=A^{+}=({{\mu }^{\prime }}_j^{-},{{\nu }^{\prime }}_j^{-},{{\pi }^{\prime }}_j^{-});j=1,2,\dots ,n$$

(4)

$$\begin{array}{l}{{\mu }^{\prime }}_j^{*}=\left\{\left(\underset{i}{\max }\left\{{{\mu }^{\prime }}_{ij}\right\}j\in J^{+}\right),\left(\underset{i}{\min }\left\{{{\mu }^{\prime }}_{ij}\right\}j\in J^{-}\right)\right\}\\ {{\nu }^{\prime }}_j^{*}=\left\{\left(\underset{i}{\min }\left\{{{\nu }^{\prime }}_{ij}\right\}j\in J^{+}\right),\left(\underset{i}{\max }\left\{{{\nu }^{\prime }}_{ij}\right\}j\in J^{-}\right)\right\}\end{array}$$

(5)

$$\begin{array}{l}{{\mu }^{\prime }}_j^{-}=\left\{\left(\underset{i}{\min }\left\{{{\mu }^{\prime }}_{ij}\right\}j\in J^{+}\right),\left(\underset{i}{\max }\left\{{{\mu }^{\prime }}_{ij}\right\}j\in J^{-}\right)\right\}\\ {{\nu }^{\prime }}_j^{-}=\left\{\left(\underset{i}{\max }\left\{{{\nu }^{\prime }}_{ij}\right\}j\in J^{+}\right),\left(\underset{i}{\min }\left\{{{\nu }^{\prime }}_{ij}\right\}j\in J^{-}\right)\right\}\end{array}$$

(6)

In this research, the first two criteria, namely, price and lead time, are classified under the cost category, while the remaining criteria fall into the profit category, please check

Table 9. A⁺ and A⁻.

Criteria	${\mathit{A}}^{+}$			${\mathit{A}}^{-}$			Criteria	${\mathit{A}}^{+}$			${\mathit{A}}^{-}$
C1	0.592	0.305	0.103	0.848	0.137	0.015	C14	0.607	0.325	0.068	0.302	0.595	0.103
C2	0.301	0.594	0.105	0.493	0.392	0.115	C15	0.188	0.747	0.064	0.073	0.865	0.062
C3	0.331	0.565	0.104	0.120	0.786	0.094	C16	0.591	0.356	0.053	0.235	0.666	0.099
C4	0.461	0.489	0.050	0.271	0.647	0.082	C17	0.519	0.363	0.118	0.254	0.644	0.102
C5	0.346	0.612	0.042	0.242	0.685	0.072	C18	0.554	0.364	0.081	0.364	0.532	0.104
C6	0.589	0.336	0.076	0.241	0.661	0.098	C19	0.477	0.472	0.050	0.190	0.731	0.079
C7	0.588	0.308	0.104	0.454	0.448	0.099	C20	0.528	0.374	0.098	0.199	0.676	0.125
C8	0.394	0.547	0.059	0.198	0.718	0.083	C21	0.408	0.491	0.101	0.180	0.729	0.091
C9	0.404	0.550	0.046	0.189	0.744	0.067	C22	0.695	0.253	0.053	0.387	0.509	0.105
C10	0.462	0.440	0.098	0.185	0.694	0.121	C23	0.853	0.134	0.013	0.371	0.530	0.099
C11	0.808	0.164	0.028	0.275	0.615	0.109	C24	0.466	0.484	0.050	0.207	0.719	0.074
C12	0.540	0.342	0.119	0.182	0.700	0.118	C25	0.802	0.168	0.030	0.379	0.523	0.098
C13	0.367	0.560	0.073	0.141	0.778	0.081	C26	0.346	0.612	0.042	0.242	0.685	0.072

.

• Step 6. Calculating the separation measures.

This research paper utilizes the normalized Euclidean distance as a means of quantifying the separation of each alternative, denoted as $S_i^{+}$ and $S_i^{-}$, from A⁺ and A⁻.

$$\begin{array}{l}{S}_i^{+}=\sqrt{\frac{1}{2n}}\sqrt{{\displaystyle \sum _{j=1}^n[({{\mu }^{\prime }}_{ij}}-{{\mu }^{\prime }}_j^{*}{)}^2+{({v^{\prime }}_{ij}-{v^{\prime }}_j^{*})}^2+{({{\pi }^{\prime }}_{ij}-{{\pi }^{\prime }}_j^{*})}^2]}\\ S_i^{-}=\sqrt{\frac{1}{2n}}\sqrt{{\displaystyle \sum _{j=1}^n[({{\mu }^{\prime }}_{ij}}-{{\mu }^{\prime }}_j^{-}{)}^2+{({v^{\prime }}_{ij}-{v^{\prime }}_j^{-})}^2+{({{\pi }^{\prime }}_{ij}-{{\pi }^{\prime }}_j^{-})}^2]}\end{array}$$

(7)

• Step 7. Calculating the relative closeness coefficient.

The computation of the relative closeness coefficient ($C_i^{*}$) for alternative Si was achieved using Equation (8). The resulting values are presented in

Table 10. S⁺, S⁻, and $C_i^{*}$ values.

Suppliers	${\mathit{S}}^{+}$	${\mathit{S}}^{-}$	${\mathit{C}}_{\mathit{i}}^{*}$	Ranking
				Intuitionistic Fuzzy TOPSIS	Fuzzy TOPSIS
S1	0.1361	0.1971	0.5915	2	2
S2	0.1957	0.1517	0.4367	3	3
S3	0.2282	0.1106	0.3264	4	4
S4	0.1192	0.2358	0.6642	1	1

. As a consequence, the best supplier in terms of LARGS criteria, S4, was identified, and the ranking of the alternatives was determined to be S4 > S1 > S2 > S3.

To demonstrate the applicability and effectiveness of the proposed methodology, we conducted a comparative analysis, using the fuzzy TOPSIS method proposed by Nazari-Shirkouhi et al. [14] and Samadi et al. for comparison [105]. The relevant results for the fuzzy TOPSIS method are shown in the last column of Table 10. In this regard, the intuitionistic fuzzy TOPSIS method effectively differentiates between the alternatives’ diversity, while there is not much difference when ranking the alternatives using the fuzzy TOPSIS method.

$$C_i^{*}=\frac{S_i^{-}}{S_i^{+}+S_i^{-}}$$

(8)

6. Managerial Implication

This study aims to introduce a supplier selection model grounded in the LARGS paradigm, encompassing lean, agile, resilient, green, and sustainable principles. The findings indicate that the model was able to successfully align with the data, demonstrating the efficacy of incorporating LARGS factors in the supplier selection process.

The results show the role of leanness in supplier selection, emphasizing that leanness should be considered in supplier selection. This finding is consistent with the results presented in [11,53,54,55,62,63,65]. This finding indicates that leanness focuses on eliminating waste, reducing inventory during manufacturing, increasing material speed in the supply chain, and decreasing preparation time in the production and transportation processes. Lean manufacturing also plays a crucial role in reducing waste in the logistics system. The cost of transportation accounts for 15–40% of the finished price of products. By decreasing inventories and transportation in the logistics system through lean suppliers, there will be a significant reduction in the final price, leading to increased competitiveness and improved business processes.

This study demonstrates the significance of supplier agility in the selection process, highlighting the importance of considering agility when choosing suppliers. This conclusion aligns with the findings presented in previous research [11,66,67,68,71,72,73,74]. Essentially, the findings emphasize that suppliers should possess the ability to promptly adapt to environmental changes, quickly recognize opportunities and threats, readily obtain necessary information from both suppliers and customers, make decisive decisions to manage environmental changes, effectively respond to market fluctuations, adjust short-term capacity as required, and modify order specifications based on customer demands.

The study’s results highlight the significance of resilience in supplier selection and emphasize the need to consider resilience when choosing a supplier. This finding aligns with previous findings presented in [15,22,62,72,75,77,78]. Based on this finding, it is crucial to select a supplier that can effectively respond to unexpected disruptions by swiftly restoring product flow, promptly recovering from disruptions, attaining a new and improved state after disruptions, adequately managing financial implications arising from potential supply chain disruptions, and maintaining optimal control over the structure and performance throughout disruptions.

The results demonstrate the importance of considering the environmental sustainability of suppliers when selecting them. This finding aligns with the results of previous studies [53,54,60,72,80,81,83,84]. It highlights the significance of managing and coordinating relationships with suppliers and incorporating environmental sustainability practices into a supply chain for long-term success. From a broader perspective, integrating environmental considerations is crucial for enhancing the capacity to develop green products and creating markets for environmentally friendly products. Implementing a green supply chain necessitates adopting new approaches, providing companies with opportunities to invest in designing and producing greener products to meet market demands. This not only involves consumer products but also encompasses collaborating with suppliers to establish green markets. Beyond cost reduction, a green supply chain facilitates the expansion of the range of available sustainable products through close collaboration with suppliers. In order to succeed, it is essential for companies to reassess their products and ensure their compatibility with the environment, establishing strong partnerships with suppliers in this process.

The findings of this study demonstrate the significance of considering environmental sustainability in supplier selection and emphasize the need to consider sustainability when choosing suppliers. These results align with the findings presented in previous research [11,79,80,83,87,88,89,90,91]. Accordingly, it is recommended that companies prioritize suppliers who actively strive to reduce energy and material consumption, focus on reusing recycled materials, utilize environmentally friendly materials in when manufacturing products, design products with standardized parts for easy reuse, facilitate product disassembly, employ life cycle analysis to assess environmental impacts, have official guidelines for sustainable product design, continuously update processes to minimize environmental impacts, and promote eco-friendly relationships in their production processes. The adoption of sustainable supply chain management offers various benefits, including environmental advantages, improvements in supply chain satisfaction, access to new markets through the supply of eco-friendly products, cost reductions achieved through resource savings, lower fuel costs, reduced working hours for employees, waste elimination, and enhanced productivity. Furthermore, implementing sustainable practices provides a competitive edge by creating value and delivering it to customers in terms of product quality, thereby leading to increased profitability.

7. Conclusions

In conclusion, this study has provided valuable insights into the significance of the LARGS paradigm in supplier selection. The findings highlight the importance of considering the lean, agile, resilient, green, and sustainable paradigm in order for companies to gain a competitive advantage. However, it is important to acknowledge the limitations of this study and consider potential hurdles with respect to implementing the proposed supplier selection model in different contexts. One limitation of this study is that the interdependencies among criteria and sub-criteria were not considered. This could have influenced the results; therefore, future research should aim to address this limitation. One possible avenue for further investigation is to assess the interdependencies among criteria and sub-criteria using the Analytic Network Process (ANP). By incorporating the ANP methodology, researchers can gain a more comprehensive understanding of the relationships among various criteria and their impacts on supplier selection outcomes. Additionally, the application of the intuitionistic fuzzy ANP method, which considers the interrelationships among criteria or alternatives, could be considered as an extension of this work. Furthermore, a sensitivity analysis assessing how changes in criteria weights impact the results can be conducted. A more extensive comparative analysis or benchmarking against industry standards or competitors would provide a broader perspective of supplier selection practices. Longitudinal studies can be undertaken to assess the long-term effectiveness of the suppliers selected using the proposed model. Additionally, integrating qualitative data into the supplier selection process by considering factors such as trust, communication, and cultural compatibility would provide a more holistic evaluation. Dynamic supplier selection models that consider changes over time should also be explored, as they can capture the dynamic nature of supplier performance. It is important to incorporate risk management considerations into the supplier selection process to ensure that potential risks are properly evaluated and managed. Lastly, a framework for evaluating the environmental impact of suppliers should be developed to promote sustainable supply chain practices. This would enable companies to prioritize suppliers based on their environmental footprints and contribute to a more sustainable business ecosystem. In summary, while this research underscores the importance of the LARGS criteria in supplier selection and presents an effective integrated approach using SEM and intuitionistic fuzzy TOPSIS analysis, there are several areas for further exploration. By considering the interdependencies among criteria and sub-criteria and incorporating the suggested avenues for future research, researchers can enhance the understanding of supplier selection processes and provide valuable insights for managers seeking to improve their supplier selection practices.