p–ISSN: 2723 - 6609 e-ISSN: 2745-5254

Vol. 5, No. 6 June 2024 http://jist.publikasiindonesia.id/

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3006

Integrated Supply Chain Quality Management and an

Organizational Performance Insights: a two-stage PLS-SEM

and Artificial Neural Network (ANN) approach

Fajar Rio Kusviansyah

, Romadhani Ardi

Universitas Indonesia Depok, Indonesia

Email: [email protected]

, [email protected]

*Correspondence

ABSTRACT

Keywords: Quality

Management, Supply

Chain Management,

Organizational

Performance, PLS-SEM,

ANN.

To generate value and optimize profitability, building

successful partnerships with supply chain organizations is

essential. This can be accomplished through training models,

knowledge transfer, and support from top management. The

implementation of advanced management practices is

crucial to achieving these goals. In this context, an integrated

approach to quality management, logistics, and supply chain

management (SCM) is fundamental. Thus, harnessing the

synergy between Quality Management (QM) and SCM is

vital to enhance and promote organizational performance.

Additionally, this study examines the significance and

relationship between knowledge transfer, supply chain

management capabilities, and top management support for

organizational performance using PLS-SEM and ANN

approaches. Primary data were collected through

questionnaires from 200 respondents working in the

manufacturing industry in Indonesia. Statistical analyses

were performed using PLS-SEM with SmartPLS 4.0, and

ANN with SPSS. The results reveal that supply chain quality

management practices and top management support

positively impact and have a strong relationship with

organizational performance. This study provides insights

into the role of supply chain quality management in

organizational performance, especially in developing

countries like Indonesia. It aims to help all manufacturing

companies enhance their organizational performance by

optimally combining selected SCQM practices with a focus

on organizational performance.

Introduction

Digital globalization has created fierce competition in today's complex business

world, causing significant barriers for organizations striving to achieve good

performance. The concept of organizational performance refers to achievements beyond

organizational profitability and places greater emphasis on economic, social, and

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3007

environmental aspects, which are increasingly demanded by various stakeholders(Arora,

Arora, Sivakumar, & Burke, 2020). With the growing concern for the environment as a

key determinant of a company's performance, it is crucial to investigate environmentally

friendly behaviour within the context of developing countries. By examining both

economic and non-economic aspects, a company's long-term performance is regarded as

the value derived from its communication principles and strategies (Ahmad, Ikram,

Rehan, & Ahmad, 2022).

Supply Chain Management (SCM) facilitates the integration of customer bases,

distribution networks, internal activities within the company, and supply bases.

Consequently, SCM practices significantly impact organizational performance and

sustainability performance, as well as how these are perceived by external stakeholders.

In the current era of globalization and increasing competition, the strategic management

of all external and internal stakeholders, from raw material suppliers to end-users, is a

primary focus of SCM. This positions SCM as a crucial management method that

influences an organization's sustainability performance (Lim, Lee, Foo, Ooi, & Wei–Han

Tan, 2022). In today's intricate economic environment, digital globalization has

intensified competition, posing serious obstacles to businesses aiming to attain

sustainability. The notion of sustainability pertains to accomplishments that surpass the

profitability of a firm and prioritizes economic, social, and environmental facets, which

are progressively required by diverse stakeholders (Bastas & Liyanage, 2019).

Supply Chain Quality Management (SCQM) is a dynamic management approach

that combines supply chain management (SCM) and quality management (QM) to

enhance customer satisfaction and superior product and service quality by fostering

cooperation between manufacturers and external stakeholders (Fernandes, Vilhena,

Oliveira, Sampaio, & Carvalho, 2022). Business organizations that successfully

implement this integrated SCQM strategy can achieve a competitive advantage and

outperform competitors in highly competitive markets. The majority of this research

reveals that SCQM can assist a business in efficiently organizing and carrying out every

activity in its supply chain, enhancing operational quality and customer satisfaction levels

(Hong, Liao, Zhang, & Yu, 2019). However, little is known about the potential

applications of SCQM, its other main drivers, and how combining these with SCQM

techniques may impact business performance.

Supply chain management practices are strongly linked to customer satisfaction. It

requires coordination and integration of the business framework. Business processes that

must be cohesive include manufacturing, purchasing, marketing logistics, and

information systems. Therefore, supply chain management practices focus on customer

response, quality, and environmental sustainability. SCM is an integrated concept that

aims to manage the upstream and downstream flow of a distribution channel from

suppliers and producers to end users (Chen, Tang, & Jia, 2019). Various SCM practices,

such as strategic partnerships with suppliers, strong relationships with customers,

maintaining the level and quality of information sharing (IS) throughout the supply chain,

and delays, occur through supply chain processes implemented by various organizations

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3008

and researchers. In contrast, QM has been defined as a critical leadership technique that

maximizes product quality, design, and features to reach customer satisfaction.

Various elements of QM have been highlighted in various studies conducted in

different sectors; however, most of the studies produced positive results for OP

(Sadikoglu & Olcay, 2014), highlighting the six total quality management (TQM) pillars

of supplier QM, knowledge and process management, leadership, training, and customer

focus (CF) all of which have been embraced by numerous other academics for use in their

empirical investigations. The literature is filled with investigations into multinational

corporations and large organizations investigating the effects of supply chain

management (SCM) and quality management (QM), two of the most significant strategic

approaches in business management that are thought to be necessary for an organization's

competitive advantage and success. Improved stakeholder satisfaction and organizational

performance are among the primary goals of both approaches. Studies on how they're

integrated into managerial techniques, however, are limited (Ahmad et al., 2022).

Furthermore, several current studies emphasize the importance of integrating SCM and

QM as a critical issue to support organizational advantages. These advantages generally

include decreased process duplication between organizations, improved departmental

coordination with a greater understanding of continuous improvement, increased process

agility in delivery, and improved customer responsiveness, all of which lead to improved

performance overall.

This work, therefore, seeks to close the gap in the existing literature by exploring

the impact of each SCQM practice on the achievement of OP among manufacturing firms

in Indonesia. A conceptual framework was developed in this study to postulate a causal

relationship between SCQM and organizational performance. This allows statistical

models to evaluate and identify SCQM factors or activities that may influence

organizational performance. Structural Equation Modelling (SEM) is used to test the

framework and implications for the success of SCQM derived from statistical

applications. An Artificial Neural Network (ANN) method will then be used which is

useful for research involving predictive settings and limited theory and does not require

an understanding of the underlying correlations between the variables under study. The

ANN method contributes to a significantly parallel distributed network consisting of

simple processing units with a tendency for the formed network to store and provide

experimental knowledge to be used as an additional analysis method to verify study

results generated by other analysis methods, such as Structural Equation Modelling

(SEM). Since top management's support and role in supply chain quality management

practices seem to be limited in prior research, it is also examined supply chain quality

management practices in organizational performance indicators. The two concepts were

integrated to comprehend the intricate behaviour of organizational tools in the supply

chain quality management implementation.

A previous study by (Ananda, Astuty, & and Nugroho, 2018) discussed the primary

SCQM dimension and the relationship between it and its impact on organizational

performance, particularly when using the Balanced Scorecard (BSC) perspective without

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3009

providing an empirical basis. A survey was conducted in 2015 regarding global

manufacturing and service companies, and statistical analysis was also conducted. The

results will help to better understand the primary issues of QM and SCM integration by

providing solid support to both practice and academics. Businesses must provide

customers with timely and efficient products and services at reasonable prices, and within

a reasonable time frame to succeed in the global marketplace. Therefore, it is imperative

to ensure quality throughout the supply chain to continuously improve quality, fulfil

customer needs and expectations, and deliver products and services of high quality. This

scenario brings the idea of integration between QM and SCM practices, stemming from

the importance of both approaches in a company to achieve competitive advantage.

(Fernandes et al., 2022) Discovered parallels between QM and SCM to shed light on the

phenomenon of SCQM integration. Consequently, the relationship between SCQM

dimensions and organizational performance indicators was explained by developing a

theoretical framework. The QM and SCM domains share many crucial areas—

sustainability, strategic management and planning, stakeholder engagement and

commitment/policy, information, leadership, continuous improvement, and innovation—

according to the SCQM conceptual model. Furthermore, it considers particular

procedures associated with SCM (procurement, internal logistics, and distribution) and

QM (product/service quality and quality culture) throughout important supply chain

processes. The integration of SCM with QM is seen as a prerequisite for achieving a

competitive advantage in the global market (Phan, Nguyen, Trieu, Nguyen, & Matsui,

2019) However, the emergence of global supply chains and the growth and expansion of

supply chain systems are forcing manufacturing companies to implement inter- and intra-

organizational management strategies (i.e., SCQM) to be sustainable in a competitive

market environment. This is because there is little research on the integration of these two

concepts. Quality has become a top priority for all parties involved in a supply chain and

a key supply chain goal. Relationships both inside and outside the organization contribute

to quality performance (Fernandes, Sampaio, Sameiro, & Truong, 2017). One of the

elements of the value-added process in the supply chain's product delivery and production

is quality (Kumar, Singh, & Modgil, 2023).

Method

This study's cross-sectional design and narrow focus on Indonesia's manufacturing

were used to test the hypotheses. The Smart-PLS (SEM) and SPSS (ANN) approaches

were also used to investigate the relationship among the variables. Other disciplines, such

as engineering, sciences, and management studies, also used this course of study.

Data Collection

To ensure that the sample chosen is a valid representation of the total population,

this study was carried out via an online survey of manufacturing enterprises located

throughout Indonesia to verify the assumptions. Top, mid, and executive-level employees

were among the industry respondents who completed this survey; they were primarily

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3010

from the departments of human resource management, corporate policy, sales, marketing,

logistics, and corporate policy.

Four experts in the field of supply chain quality management reviewed the

questionnaire's content validity before distribution; as a result, several modifications were

made to prevent future misunderstandings and better suit the Indonesian environment. To

evaluate the general design of the questionnaire and the comprehensiveness of the scale

items, a pilot test involving 40 participants was carried out. Moreover, the pilot study was

conducted to rigorously validate the instruments utilized in this study. In the pilot study,

Cronbach’s alpha of all the constructs was greater than 0.7, which refers to high internal

consistency. Based on the research's suggested assumptions, a questionnaire was created

to gather information from Indonesia's industry regarding supply chain quality control

and disposal decisions. The survey was modified based on (Fernandes et al., 2022). Data

from Indonesia's industrial industry was gathered online using Google and a self-

administered questionnaire. The respondents' email addresses received a cover letter and

the study link. Web-based surveys are becoming increasingly popular among researchers

because they save time, are more easily accessible, cost-effective, and provide real-time

data.

A total of 274 completed questionnaires were collected. However, 74

questionnaires were incomplete, so only 200 samples could be used. Using the P-min

introduced by (Hair & Alamer, 2022), The study required a minimum of 155 respondents,

and since 200 questionnaires were completed, the sample size was deemed acceptable

and compliant.

Data Analysis Technique

The information was collected and numerically coded in an MS Excel spreadsheet

before being transferred to the Smart-PLS and SPSS software for further analysis and

assessment. The data was visually checked for errors and missing numbers to make the

necessary adjustments. The structural equation modelling technique was used to analyze

both the measurement and structural models. Furthermore, utilizing the SEM analysis, it

was discovered that structural equation modelling could assess the measurement models

(Hair & Alamer, 2022). Using latent components, causal models with their observable

variables can be systematically analyzed. This study adopted a two-step methodological

approach in which the measurement model was analyzed initially, followed by the

estimation of structural equation modelling. Smart-PLS software was used to analyze

both the measurement and structural equation modelling. To complement PLS-SEM

analysis, which was applied to discover statistically significant reflective independent

variables that affect a reflective dependent variable, artificial neural network (ANN)

analysis was subsequently carried out using SPSS. These significant reflective

autonomous variables are then treated as input neurons for the resultant output neuron

(the reflective dependent variable) in the ANN analysis. Because of its flexibility and lack

of reliance on multivariate assumptions like normality and linearity, ANN is comparable

to PLS-SEM.

Demographic profile of respondents

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3011

The full demographic proﬁle of the respondents is shown in Table 1. From the

gender proﬁle, there were more male (57.5%) than female (42.5%) respondents. The

primary activities of 22% of the companies classified as "Other" include the production

of solar panels, medical equipment, solder products, cement products, and

remanufacturing of recyclable products, and so fort second of primary manufacture

automotive is 20%, food and beverage 19,5%, textile 12% and last is pharmaceutical

10,5%. For the division, most of the respondents work in the production division with

16%, the second is procurement and quality assurance with 11,5%, and the last is

sales/marketing with 6%. Additionally, 79% of the respondents held junior management

positions (supervisor and assistant manager), while only 21% held top management

positions.

Results and Discussion

Measurement Models

To ensure that the measured variables were accurately observed, the validity and

reliability of the suggested model frameworks were examined in the measuring model.

Using Smart PLS software, the item's internal consistency, convergent and discriminating

validity, and reliability were assessed. Moreover, the individual dependability of each

product was assessed to determine whether a latent variable accounts for a significant

percentage of the variance in its observed indices. It was determined by looking at the

loading with each instrument's structure, and it needs to be at least 0.5. Internal

consistency was also examined. This indicator measures the constructed concept while

testing the internal consistency of all indicators. Internal consistency was also examined.

This indicator measures the constructed concept while testing the internal consistency of

all indicators. It demonstrates the rigour with which the same latent variable is measured

by the deeply embedded elements. It can be assessed using composite reliability or

Cronbach's alpha. According to (J. Hair & Alamer, 2022), Cronbach's alpha indicates that

results have internal consistency reliability and that all loading variables are equally

correct. The internal consistency of the recommended model under evaluation should be

at least 0.7 in terms of composite reliability. The assessment model includes two parts:

convergence validity and decrement validity. The assessment of convergence validity

involved factor loadings (FL), composite reliability (CR), and average variance extracted

(AVE). The outcomes show that most factor loadings for items exceeded the minimum

value of 0.7 as Table 2 displays the result.

Table 1

Measurement Model

Constructs

Factor Loading

(FL)

Composite

Reliability

(CR)

Average

Variance

Extracted

(AVE)

Organizational

Performance

0.935

0,645

OP 1

0.883

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3012

OP 2

0.836

OP 3

0.808

OP 4

0.767

OP 5

0.724

OP 6

0.81

OP 7

0.778

OP 8

0.808

Supply Chain

Management

0.912

0.676

SCM 1

0.851

SCM 2

0.763

SCM 3

0.91

SCM 4

0.825

SCM 5

0.753

Quality

Management

0.924

0.708

QM 1

0,824

QM 2

0.821

QM 3

0.847

QM 4

0.836

QM 5

0.876

Supply Chain

Quality

Management

Practices

0.897

0.743

SCQMP 1

0.869

SCQMP 2

0.814

SCQMP 3

0.901

Supply Chain

Quality

Management

Capabilities

0.916

0.732

SCQMC 1

0.869

SCQMC 2

0.814

SCQMC 3

0.901

SCQMC 4

0.916

Support Top

Management

0.906

0.763

STM 1

0.771

STM 2

0.926

STM 3

0.915

Knowledge

Transfer

0.90

0.645

KT 1

0.717

KT 2

0.789

KT 3

0.853

KT 4

0.84

KT 5

0.808

Additionally, the constructs' CA and CR values were above 0.7, with AVE values

surpassing the threshold of 0.5. Thus, these latent variables have no issues regarding

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3013

convergent validity. The Heterotrait- Monotrait was used to test the discriminant validity

model. Table 3 displays the results, which demonstrate that the discriminant values were

within the HTMT threshold of 85 or 90, confirming the presence of discriminant validity.

Table 2

Discriminant validity based on the HTMT method

SCM

SCQMC

SCQMP

STM

0.586

SCM

0.514

0.588

SCQMC

0.691

0.791

0.897

SCQMP

0.646

0.814

0.672

0.888

STM

0.547

0.805

0.485

0.674

0.828

0.379

0.483

0.787

0.824

0.637

0.55

Note : OP =Organitazional Performance ; QM = Quality Management ; SCM = Supply

Chain Management ; SCQMC: Supply Chain Quality Management Capabilities ; SCQMP

= Supply Chain Quality Management Practices ; STM= Support Top Management; KT=

Knowledge Transfer

Structural Model

The structural and measurement models of this study are based on a conceptual

model consisting of latent variables and indicators. Figure 2 shows the structural and

measurement models of this study.

Figure 1 Develop PLS-SEM Structural Path Analysis for Factors Affecting OP in SCQM.

The hypotheses were tested using PLS-SEM, which followed the validation of the

measurement model. As suggested by Hair et al., (2019), the reporting includes path-

coefficient results, inner VIF values, and coefficient of determination (R2). The

standardized R2 value ranges from 0 to + 1 with greater predictive accuracy values. Also,

if the independent variable has R2 values greater than 0.7, it is considered a strong

coefficient determinant. The variables having a value of less than 0.25 are considered

weak, and 0.5 is assumed to be moderate. Effect sizes (f2) and predictive relevance of Q2

The model is considered good if its predictive relevance in Q2 is greater than zero. The

collinearity issues have been examined to ensure that they did not affect the regression

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3014

results. According to (Hair & Alamer, 2022), the VIF inner values should be below 5.

The results indicate that all the items pass the threshold of 5. Furthermore, the tolerance

values were within the acceptable range (0.1 and 1). Hence, it is evident that the dataset

does not exhibit multicollinearity. Therefore, the bootstrapping approach with a

resampling of 1,000 was utilized to estimate the significance of the path coefficient using

SmartPLS 4 software.

Evaluation of the structural model follows the assessment of the measurement

model. The structural model was utilized to predict the underlying hypothesized

relationships among latent variables for various dimensions. Predicted construct

associations were computed using the structural path model's coefficients. The intensity

of the relationship between the two paradigms was represented by the path coefficient's

value. Its regular value fluctuates between -1 and +1. The values closer to +1 are

considered statistically significant and show a strong correlation between the constructs.

On the other hand, weak correlations are represented by values that are closer to 0, and

they are usually not significant. To evaluate the relevance of various connections between

the components in further detail. In this investigation, the crucial value was determined

using 5% of the significance level. Moreover, if the t-value is greater than the critical t-

value, the hypothesis is accepted at the mentioned significance level; otherwise, it is

rejected. Table 4 shows the values based on the evaluation criteria;

Table 3

Assessment of Structural Model

Construct

Path

Coeffici

ent

VIF

R-

Squar

)

Effe

Size

)

Q-

Squ

are

)

P -

Value

T-

Valu

Result

QM ->

SCM

0.737

1.00

0,541

0.54

0.53

19.07

0.000

QM ->

SCQMP

0.285

2.52

0.623

0.27

0.45

3.288

0.001

SCM ->

SCQMP

0.603

2.64

0.623

0.36

0.45

7.134

0.000

SCQMP ->

SCQMC

0.783

1.00

0.611

0.61

0.25

26.95

0.000

SCQMP ->

0.657

1.00

0.509

0.43

0.19

9.997

0.000

SCQMC ->

0.41

1.98

0.344

0.16

0.28

7.281

0.000

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3015

KT -> OP

-0.094

2.13

0.344

0.00

0.28

1.855

0.064

Rejecte

STM ->

SCQMP

-0.056

2.35

0.623

0.00

0.45

6.073

0.031

STM ->

0.375

1.30

0.344

0.14

0.28

7.091

Note : OP =Organitazional Performance ; QM = Quality Management ; SCM = Supply

Chain Management ; SCQMC: Supply Chain Quality Management Capabilities ; SCQMP

= Supply Chain Quality Management Practices ; STM= Support Top Management; KT=

Knowledge Transfer

Based on Table 3, the path coefficient indicates that the supply chain quality

management practices (SCQMP) variable and supply chain quality management

capabilities (SCQMC) variable have a strong positive relationship with a path coefficient

of 0.783. Subsequently, the variable quality management (QM) was influenced by supply

chain management (SCM) with a value of 0.735; supply chain quality management

practices (SCQMP) about knowledge transfer (KT) with a value of 0.657; supply chain

management (SCM) about supply chain quality management practices (SCQMP) with a

value of 0.603; supply chain quality management capabilities about organizational

performance (OP) with a value of 0.41, support top management (STM) about

organizational performance (OP) with a value of 0,375; and quality management (QM)

about supply chain quality management practices (SCQMP) with a value of 0.285. There

are two negative correlation variables, which are support top management (STM) with

supply chain quality management practices (SCQMP) at -0.056 and knowledge transfer

(KT) with organizational performance (OP)

The structural model will be assessed using a predetermined coefficient of

determination (R2). According to Hair & Alamer, (2022), Substantial criteria were

obtained when the R2 values were 0.75 (strong), 0.50 (short), and 0.25 (long). Table 4

displays the results, which show that the R2 value for the supply chain management

variable is 0.541, or 54,1%; the supply chain quality management practices and

capabilities are 62,3% and 61,1%; the transfer of knowledge is 50,9%; and the

organizational performance is 34,4%. This indicates that the variable supply chain

management may be explained by the variable quality management, which is 54.1%.

However, supply chain management methods can be explained by the variables quality

management, supply chain management, and support top management, which account for

62,3% of the variance. In addition, practices for supply chain quality management can

explain the variable of supply chain quality management skills, which is 61.1%. Then,

the supply chain quality management practices variable can explain knowledge transfer

by 50.9%. The last variable, supply chain quality management capabilities, can explain

organizational performance by 34.4% by supporting top management and transferring

information. In this case, all of the R2 values had values that were less than 0.50, meaning

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3016

that they were classified as negative values. Only one R2 value was classified as positive

since it had a value of less than 0.50.

Another structural model evaluation method involves identifying the effect size or

(f2) which shows the contributions of each exogenous variable to the endogenous

variable. According to Leong et al., (2020b) The contribution is higher if the score f2¬¬

≥ 0,35, normal if the score f¬2 ¬¬≥ 0,15, and low if the score f¬2 ¬¬≤ 0,02. The variables

supply chain quality management capabilities (0,613) and top management support

(0,432) significantly affect organizational performance. Therefore, the supply chain

quality management technique significantly impacts knowledge transfer (0,364) and

supply chain quality management competencies (0,543). The last testing of the inner

model was conducted by examining the Q2 values using the blindfolding technique,

which assesses predictive relevance. A hypothesized model will have good predictive

power for endogenous constructs if the Q2 values exceed zero. Table 4 shows that all Q2

values are greater than 0, indicating the model has predictive relevance for the related

endogenous variables.

Finally, when assessing the effect of the exogenous variable on the significant

endogenous variable, the absolute t-statistics >1,96 and the p-value < 0.05 may be

observed. Table 4 illustrates that eight hypotheses are accepted and satisfy the criteria,

and one hypothesis is rejected because it does not satisfy the criteria.

Artificial Neural Network

The findings clearly show that organizational performance has been enhanced, and

the understanding of supply chain quality management is strengthened by having defined

capabilities and support from top management. However, non-linear linkages are beyond

the scope of this approach. To verify the PLS-SEM results and identify the non-linear

correlations, artificial intelligence analysis was used, as will be explained in the following

discussion. PLS-SEM and ANN are used to demonstrate that the relationships between

the constraints are not compensatory nor linear. Furthermore, according to Sternad

(Sternad Zabukovšek, Kalinic, Bobek, & Tominc, 2019), ANNs can be helpful in research

with little theory and a predictive setting and when comprehension of the underlying

relationships between the variables being researched is not necessary. An enormously

parallel distributed network made up of basic processing units with a neural tendency to

store and make use of experimental information is facilitated by ANN techniques.

Because of this, ANNs are frequently used as an extra analytical technique to confirm the

conclusions of research conducted using other analytical techniques, including structural

equation modelling (SEM) (Sternad Zabukovšek et al., 2019). In the most recent study,

Sternad Zabukovšek et al., (2019) found that combining the ANN and SEM methods

yields a more thorough analysis that helps to assess the relationships between each

predictor more precisely. Moreover, by identifying both linear and nonlinear correlations

between every variable, ANN yields extremely accurate results.

Given the contributions and acceptability of the ANN approach for evaluating our

ﬁndings, we use ANN to measure the relationship between each predictor (i.e. SCQMC,

KT, and STM) and the dependent variable (i.e. OP). SPSS was used to conduct ANN

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3017

analysis. The model architecture uses the sigmoid function as the activation function in

the hidden and output layers. In addition, to increase the effectiveness of training and

obtain better model performance, all inputs and outputs are normalized to the range [0,

1]. The number of neurons in the input layer is equal to the number of predictor variables.

In contrast, the number of neurons in the output layer is equal to the number of dependent

variables, i.e., prediction variables, and both are determined by the problem structure.

Furthermore, the model architecture in this study has three types of models: models with

output variables of supply chain quality management capabilities (SCQMC), knowledge

transfer (KT), and support top management (STM), with different input variables as

shown in Figure 2.

Figure 3. ANN Model Developed

All variables were re-analyzed using the ANN method because they qualified as

significant variables in the SEM results. In this case, the input data for ANN is the latent

score variable, the result of SEM analysis. The data normalization process is done

automatically by selecting the rescaling method, which is normalized. The process result

aims for normalized data to be in the range of 0 to 1. After normalizing the data, the data

partition process is carried out. To prevent over-fitting, a ten-fold cross-validation

procedure is performed, with 90 of the samples used for training and the remaining 10

used for testing.

According to Leong et al., (2015), ANN and PLS-SEM share similarities as both

are adaptable and do not require meeting multivariate assumptions such as normality and

linearity. The present study used a multilayer perceptron with a "feed-forward back-

propagation" algorithm, using the significant predictors from PLS path analysis as input

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3018

neurons. It was used because it is considered the most accepted and common model for

this type of research. The output and the hidden layer were both activated using a sigmoid

function. The RMSE values multiplied by ten are shown in Table 4.

Table 5

Artificial Neural Network values

Input: Supply chain quality management capabilities; knowledge

transfer; support top management

Output: Organizational performance

Training

Testing

SSE

RMSE

SSE

RMSE

142

36.300

0.5346

12.802

0.4915

142

34.136

0.5085

28.938

0.7765

141

36.731

0.5295

12.280

0.5006

141

36.905

0.5308

13.302

0.5210

144

37.975

0.5323

12.366

0.5185

139

39.024

0.5500

30.447

0.7727

125

35.112

0.5526

12.579

0.4399

132

36.104

0.5440

13.627

0.4847

138

34.608

0.5200

26.728

0.7169

144

36.846

0.5244

11.486

0.4997

Mean

36.374

0.533

Mean

17.456

0.572

Standard

Deviation

1.491

0.014

Standard

Deviation

7.834

0.129

In addition, the root mean square of error (RMSE) for each of the 10 neural

networks in Table 5 is computed to assess the predictive accuracy of the ANN model.

Given that a lower RMSE value denotes improved data fitness and increased prediction

accuracy. The ANN model developed in this study shows higher prediction accuracy and

reliable data ﬁtness because the RMSE values for training (0.533) and testing (0.572) are

relatively small.

The importance of each predictor measures how much the network model's

predicted value changes for different predictor values. The normalized or relative

importance value is calculated by dividing the importance value by the largest importance

value, and the result is expressed in percentage. Table 6 shows the results of the sensitivity

analysis calculation of the two models, namely models A and B.

Table 5

Sensitivity analysis

Neural Network (NN)

Model A (output neuron: BI)

SQCMC

STM

NN (1)

0.377

1.000

0.945

NN (2)

0.384

0.918

1.000

NN (3)

0.373

1.000

0.981

NN (4)

0.460

0.922

1.000

NN (5)

0.360

0.970

1.000

NN (6)

0.309

0.978

1.000

NN (7)

0.340

0.869

1.000

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3019

NN (8)

0.356

1.000

0.893

NN (9)

0.464

0.998

1.000

NN (10)

0.402

0.807

1.000

Average Normal

Importance

0.383

0.946

0.982

Normalized Importance

39%

96%

100%

Note : SCQMC: Supply Chain Quality Management Capabilities; STM= Support Top

Management; KT= Knowledge Transfer

The normalized relevance value was used to rank the relative significance of each

predictor variable derived from the sensitivity analysis displayed in Table 6. According

to the results of the sensitivity analysis, STM has the highest normalized relative

relevance (100%) impact on OP, followed by SCQMC (96%), and KT (39%).

This study aims to experimentally investigate whether extra STM and SCQM

practices (QM, SCM, SCQMP, SCQMC, KT) impact OP achievement. Our results

advance knowledge transmission based on capabilities theory and dynamic practice, even

though only partial support was found for the correlations between the selected predictors

and dependent variables. The knowledge-based view holds that organizations are

information-bearing entities and that a company's primary function is to integrate and

apply knowledge. Dynamic capability and practices, on the other hand, pertain to the

capacity to build, create, and integrate information resources to perceive, identify, and

address environmental difficulties. The fundamental mechanism of knowledge-based

dynamic capabilities is made up of internal and external knowledge-related activities

rooted in partnerships and networks, and our results indicate that SCQMC and STM relate

to OP positively and signiﬁcantly (i.e. normalized relative importance exceeding 50%),

where as KT was surprisingly found to have no signiﬁcant relationship to organizational

performance.

Table 6

Comparison between PLS-SEM and ANN findings

Variable

PLS SEM

ANN

Matc

hed

Not

Total

Effect

Performance

Findings

Ranking

(PLS-

SEM)

based on

Path

Coefficient

Ranking

(ANN)

[based on

Normalized

relative

Importance

(%)

ANN

results:

Normalized

relative

importance

(%)

SCQMC

0,406

57,635

Significant

96%

Not

STM

0.360

53,176

Significant

100%

Not

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3020

0.089

47,121

Not

Significant

39%

Match

The significance of the predictors based on PLS-SEM and ANN is displayed in

Table 7. Each variable's Path Coefficient and Normalized Relative Importance statistics

affect its ranking in both PLS-SEM and ANN. The fascinating results originate from

SQCMC's first-place rating in PLS-SEM influence rankings; conversely, it ranks second

in ANN rankings. On the other hand, STM is ranked first in ANN and second in PLS-

SEM in terms of influence. Still, KT comes in third in both categories. This discrepancy

could be explained by ANN's unique ability to capture both linear and non-linear

relationships. The data unequivocally shows that when the non-linear relationship with

OP is considered, the significance of STM increases. Using only PLS-SEM would have

obscured this crucial insight. The combined use of PLS-SEM and ANN techniques in our

study offers a comprehensive perspective, enhancing the robustness and validity of our

findings. Many studies have employed either technique in isolation, but our integrated

approach provides a richer, more nuanced understanding of the relationships among the

variables.

Theoretical Implications

The most crucial SCQM practice for improving an OP, practically speaking, was

determined to be SCQMC. Therefore, to achieve customer satisfaction and OP,

manufacturing firms should focus on sustaining tight cooperation and partnerships with

their consumers. A company can accomplish quality objectives and advance desired

sustainability results by working cooperatively with its consumers. Furthermore, our

results demonstrate that STM is crucial to manufacturing firms' ability to maintain

performance. To achieve the highest quality outcomes and the desired OP, Senior

management support and initiatives to foster a work environment open to continuous

improvement are therefore crucial. This study demonstrates that superior management's

ability to influence people and manage resources in manufacturing companies has aided

in achieving OP and improved quality of life.

The primary conclusion of the present study, concerning the relationships, is that

KT has no discernible influence on the linkages between SCQMP and organizational

performance. This outcome does not align with the findings. This demonstrates how

successfully implementing the SCQM agenda may enhance businesses' knowledge

transfer (KT) procedures, which has a major impact on business performance.

Additionally, techniques for exchanging expertise with distributors offer a chance to

obtain a competitive edge. Top management should set the stage for the formation and

exchange of knowledge inside the supply chain by providing the required tools, operating

systems, frameworks, and protocols. The OP of the bigger organization can be a source

of knowledge transfer.

Manufacturers are thought to resist achieving OP since integrating the whole supply

chain takes a significant amount of time and money. Consequently, it has been discovered

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3021

that SCQM methods like QM, SCM, SCQMP, and SCQMC significantly affect

manufacturers in developing nations like Indonesia. By combining specific SCQM

methods with an emphasis on organization, Indonesian manufacturing companies may be

able to improve their OP using the study's findings. Manufacturers can then choose which

area to focus on to find effective and successful solutions to issues brought about by

careless employees who frequently neglect SCQM standard practices.

Managerial Implications

Our study model has the potential to be a useful tool for future studies on

sustainability. Additionally, the study model expands on recent SCM and QM literature.

Foo et al., (2018 ) and Soares et al., (2017) examined each of the chosen SCQM practices

separately on the three OP aspects. As a result, manufacturing companies can prioritize

SCQM procedures that capitalize on their sustainable outcomes. Furthermore, the study's

theoretical framework for SCQM procedures invites researchers to use and build upon

these data in further research. As a modern sustainability approach that entails embracing

a new management strategy to tackle dynamic organizational issues by incorporating the

TBL into organizational management practices, the theoretical contribution of SCQM is

linked to the philosophy of QM and SCM. Moreover, Indonesian OP responsiveness is

still developing. Manufacturing companies appear to be exempt from significant

environmental regulations, which have produced enormous amounts of waste, the

depletion of natural resources, and the excessive use of energy. Environmental laws about

the manufacturing sector must be implemented and strictly enforced to solve issues with

waste and pollution. Thus, by employing an empirical methodology, this study

contributes to sustainability research by highlighting the benefits of applying SCQM

techniques to achieve sustainability goals, particularly in OP research. Furthermore, this

study establishes the groundwork for future empirical investigations into the feasibility

of applying the suggested research paradigm, in whole or in part, to another nation.

Finding out how the research model functions in international commercial enterprises will

be fascinating. The goal of this study is to overcome businesses' ignorance of open-poly

and persuade them of the benefits of open-poly in terms of long-term success.

Consequently, businesses can acquire positive approaches and modify their immediate

negative.

Conclusion

In conclusion, this study looks at the range of SCQM procedures that manufacturing

companies considering SP might implement. Many sustainable result studies have

considered both SCM and QM practices; however, the current literature does not present

the convergence of these two techniques into a set of SCQM practices, as this study does.

Moreover, this study can aid in the successful application of SCQM practices since it

integrates behavioural or soft (i.e., SCQMC, STM, and TK) and technical or hard (i.e.,

SCQMC) views of SCQM practices. According to (Kumar et al., 2023), firms in emerging

economies must adopt behavioural or soft practices of SCQM to achieve OP. This is

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3022

because soft practices can foster an environment in which hard SCQM practices can be

further developed by the firms. In response to all hypotheses, the study's findings

highlight the significance of the behavioural components of OP that are SCQMC,

particularly for STM, which has the biggest influence on SCQMC and KT, come in

second and third, respectively, according to our ANN sensitivity analysis. ANN shows

that STM has the greatest impact on OP among the significant predictors identified,

whereas KT has the least.

Integrated Supply Chain Quality Management and an Organizational Performance Insights: a

two-stage PLS-SEM and Artificial Neural Network (ANN) approach

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3023

Bibliography

Ahmad, Azhar, Ikram, Amir, Rehan, Muhammad Farooq, & Ahmad, Ayyaz. (2022).

Going green: Impact of green supply chain management practices on sustainability

performance. Frontiers in Psychology, 13, 973676.

Ananda, Armalia Reny Wijayanti, Astuty, Pudji, & Nugroho, Yudhinanto Cahyo. (2018).

Role of green supply chain management in embolden competitiveness and

performance: Evidence from Indonesian organizations. International Journal of

Supply Chain Management, 7(5), 437.

Arora, Amit, Arora, Anshu Saxena, Sivakumar, K., & Burke, Gerard. (2020). Strategic

sustainable purchasing, environmental collaboration, and organizational

sustainability performance: the moderating role of supply base size. Supply Chain

Management: An International Journal, 25(6), 709–728.

Bastas, Ali, & Liyanage, Kapila. (2019). Integrated quality and supply chain management

business diagnostics for organizational sustainability improvement. Sustainable

Production and Consumption, 17, 11–30.

Chen, Lujie, Tang, Ou, & Jia, Fu. (2019). The moderating role of supplier involvement

in achieving sustainability. Journal of Cleaner Production, 235, 245–258.

Fernandes, Ana Cristina, Sampaio, Paulo, Sameiro, Maria, & Truong, Huy Quang.

(2017). Supply chain management and quality management integration: A

conceptual model proposal. International Journal of Quality & Reliability

Management, 34(1), 53–67.

Fernandes, Ana Cristina, Vilhena, Estela, Oliveira, Rui, Sampaio, Paulo, & Carvalho,

Maria Sameiro. (2022). Supply chain quality management impact on organization

performance: results from an international survey. International Journal of Quality

& Reliability Management, 39(2), 630–646.

Hair, Joseph, & Alamer, Abdullah. (2022). Partial Least Squares Structural Equation

Modeling (PLS-SEM) in second language and education research: Guidelines using

an applied example. Research Methods in Applied Linguistics, 1(3), 100027.

Hong, Jiangtao, Liao, Yi, Zhang, Yibin, & Yu, Zhefu. (2019). The effect of supply chain

quality management practices and capabilities on operational and innovation

performance: Evidence from Chinese manufacturers. International Journal of

Production Economics, 212, 227–235.

Kumar, Anil, Singh, Rohit Kumar, & Modgil, Sachin. (2023). Influence of data-driven

supply chain quality management on organizational performance: evidence from

the retail industry. The TQM Journal, 35(1), 24–50.

Lim, Ai Fen, Lee, Voon Hsien, Foo, Pik Yin, Ooi, Keng Boon, & Wei–Han Tan, Garry.

(2022). Unfolding the impact of supply chain quality management practices on

Fajar Rio Kusviansyah, Romadhani Ardi

Jurnal Indonesia Sosial Teknologi, Vol. 5, No. 6, June 2024 3024

sustainability performance: an artificial neural network approach. Supply Chain

Management: An International Journal, 27(5), 611–624.

Phan, Anh Chi, Nguyen, Hao Anh, Trieu, Phuong Dinh, Nguyen, Ha Thu, & Matsui,

Yoshiki. (2019). Impact of supply chain quality management practices on

operational performance: empirical evidence from manufacturing companies in

Vietnam. Supply Chain Management: An International Journal, 24(6), 855–871.

Sadikoglu, Esin, & Olcay, Hilal. (2014). The effects of total quality management practices

on performance and the reasons for and the barriers to TQM practices in Turkey.

Advances in Decision Sciences, 2014.

Sternad Zabukovšek, Simona, Kalinic, Zoran, Bobek, Samo, & Tominc, Polona. (2019).

SEM–ANN-based research of factors’ impact on extended use of ERP systems.

Central European Journal of Operations Research, 27, 703–735.