Construction of Evaluation Index System of Harmonious Labor Relations in Chinese Enterprises: An Empirical Analysis Based on Structural Equation Model

Abstract

Since entering the period of economic and social transformation, China’s labor relations conflicts have entered a period of prominence and prevalence, which has brought great challenges to ensure the sustainable development of enterprises and build a harmonious society. Constructing a scientific and reasonable evaluation index system of harmonious labor relations in enterprises to measure the harmonious degree of labor relations is of great significance for improving the management level of labor relations in enterprises, promoting sustainable development of enterprises, and building a socialist harmonious society. Based on the relevant research results of labor relations and from the perspective of employees, this paper took the ERG theory (E stands for existence needs, R stands for relatedness needs, G stands for growth needs) as the basic framework of the evaluation index system, combined with the literature research method and expert consultation method to initially design evaluation index system, used SPSS 26.0 software and principal component analysis to further determine the evaluation index system structure. Based on the structural equation model, this paper used AMOS 23.0 software to carry out first-order and second-order confirmatory factor analysis for the evaluation index system and assigned objective weights to the evaluation indicators according to standardized factor analysis path coefficients. The index system, via structure confirmation and index empowerment, is used to evaluate the harmonious degree of labor relations between the two enterprises, N and Z. It is found that the evaluation index system of enterprise harmonious labor relations constructed in this paper conforms to the theoretical hypothesis model; The index system includes salary and benefits, working conditions, communication management, interpersonal relationship, individual participation and personal development six dimension-level indexes and 18 index-level indexes; The overall labor relations of companies N and Z are relatively harmonious, but the relative satisfaction of individual development indicators at the dimension level is low, and there is still room for improvement. The research results of this paper not only enrich and improve the theoretical system of labor relations to a certain extent but also provide a new perspective for the integration of enterprise harmonious labor relations and ERG theory and provide a theoretical basis and practical guidance for enterprises to establish harmonious labor relations and promote sustainable development of enterprises.

1. Introduction

Labor relations, as the most fundamental social relations, are the basis of the whole of social and economic life. Harmonious labor relations are not only conducive to the sustainable development of enterprises but also an important part of building a harmonious socialist society. Harmonious labor relations can not only improve the satisfaction and loyalty of employees maintain the stability of enterprises, but also enhance the competitiveness and brand influence of enterprises, and promote the economic growth and sustainable development of enterprises. On 21 March 2015, the Central Committee of the Communist Party of China and the State Council issued the “Opinions on Building Harmonious Labor Relations”, which is a programmatic document guiding the construction of harmonious labor relations with Chinese characteristics during the period of economic and social transformation, and clearly defines the great significance, guiding ideology, working principles, objectives and tasks, and policy measures for building harmonious labor relations. On 3 January 2023, the National Tripartite Committee for Coordinating Labor Relations issued the “Opinions on Promoting the Establishment of Harmonious Labor Relations in the New Era” (hereinafter referred to as the “Opinions”), laying a solid foundation for further building harmonious labor relations with Chinese characteristics. The “Opinions” emphasize the principle of improving people’s well-being, clarify the people-oriented approach and the main position of employees in building harmonious labor relations, focus on promoting enterprises to respect labor and benefit employees, and stress establishing a standardized, systematic, and humanistic evaluation mechanism.

The harmonious degree of labor relations in enterprises is related to the personal interests of employees and the sustainable development of enterprises and even affects the economic development of the country and the harmony and stability of society. In the process of China’s economy shifting from the stage of high-speed growth to the stage of high-quality development, influenced by factors such as new subjects and new technologies in labor relations, a new situation has gradually emerged in the field of labor relations, and the contradictions and conflicts in China’s labor relations have gradually expanded and complicated, posing great challenges to the construction of a harmonious society [1]. On the one hand, the quality of modern talent has gradually improved, the concept of employment has changed, and the awareness of rights protection for their own labor activities has become stronger and stronger; on the other hand, due to the continuous development of new media and information technology, modern talents can spread and receive information without the constraints of time and space, to a certain extent, promote the occurrence of collective interest appeal actions [2,3]. Therefore, studying how to achieve a high degree of harmony in enterprise labor relations and constructing a scientific and systematic evaluation index system for harmonious labor relations in enterprises during the period of economic and social transformation is a good inspiration and guidance for the realistic difficulties faced by enterprises in building harmonious labor relations.

At present, Chinese and foreign scholars have carried out extensive and in-depth discussions on harmonious labor relations. Some scholars have carried out theoretical studies on the connotation, influencing factors, construction mechanism, and improvement suggestions of harmonious labor relations [4,5,6,7,8,9], while others have tried to carry out empirical studies by constructing an evaluation index system. Ma Cuihua and Kong Fanyi [10] designed the evaluation index system of harmonious labor relations in enterprises from three aspects: material culture, institutional culture, and spiritual culture, based on the perspective of cultural identity. Based on a new perspective of generalized ecosystem, Li Yulong and Song Yajie [11] constructed an evaluation index system of harmonious labor relations from five aspects: organizational culture, institutional arrangement, social economy, individual factors, and labor-capital relations; Based on the perspective of partnership, Zuo Jing, Wang Decai and Feng Junwen [12] constructed an evaluation index system for harmonious labor relations from four aspects: employee participation, work motivation, communication and development, and employment security. However, the existing research has not yet formed a unified standard for the dimension composition of the evaluation index system of harmonious labor relations in enterprises. The selection of indicators ignores the principle of humanistic needs, and few scholars construct the evaluation index system of harmonious labor relations from the perspective of employees, and empirical studies are relatively lacking. Therefore, from the perspective of employees, this paper constructs an evaluation index system of enterprise harmonious labor relations based on the new humanistic needs theory—ERG theory (E stands for existence needs, R stands for relatedness needs, G stands for growth needs), carries out an empirical analysis of enterprise harmonious labor relations evaluation index system based on structural equation model and evaluates the degree of labor relations harmony of N and Z enterprises. This paper attempts to make the following contributions to the existing literature. Firstly, this study is a beneficial exploration of harmonious labor relations in enterprises based on the perspective of employees, which enriches the research perspective of the theoretical system of labor relations and plays a certain role in promoting the development of relevant theories. Secondly, this study combines ERG theory to build an evaluation index system, which is an extension of demand theory in the evaluation of enterprise harmonious labor relations. Thirdly, the structural equation model is used to carry out first-order and second-order confirmatory factor analysis, and the standardized factor analysis path coefficient is used to objectively assign weight to the evaluation indicators so as to ensure the objective, scientific, and reasonable structure and weight of the index system. Fourth, the evaluation index system of harmonious labor relations in enterprises is constructed to achieve quantitative measurement of the harmonious degree of labor relations in actual enterprises and provide theoretical basis and practical guidance for enterprises to establish harmonious labor relations and promote sustainable development of enterprises.

2. Literature Review

2.1. Labor Relations Related Research

Foreign scholars are divided into four schools of thought around the labor relations model, namely neoliberal egoists, monists, pluralists, and critics [13]; the assumptions of each school on labor relations models cover labor relations objectives, individual behaviors and structural contexts, where structural contexts include labor market equality or not, and compatibility of interests between firms and workers or not [14]; some scholars also study labor relations in the global economy from the perspective of value chain allocation based on the global value chain [15]. Chinese scholars mainly carry out relevant research on labor relations from the perspectives of economics, public management, and business management. From the perspective of economics, Huang Weide and Chen Xin [16] divided the attitudes and behaviors of enterprises toward employees into two levels, material and spiritual, which were used to measure the level of harmonious labor relations in enterprises, thus forming a non-differential curve model of harmonious labor relations; using Pareto efficiency as a reference; Lv Jingchun and Li Yongjie [17] deduced that the base point of harmonious labor relations is to ensure a balance between the efficiency needs of employers and the equity needs of employees; He Aiping and Xu Yan [18] hold that in the process of promoting the development of labor data digitization, corresponding measures should be actively taken to establish and improve the coping mechanism, so as to promote the coordinated development of labor data digitization and labor relations. From the perspective of public management, Yang Chengxiang and Jiang Ying et al. [19,20] summarized harmonious labor relations with Chinese characteristics as standardized, orderly, fair and reasonable, mutually beneficial, harmonious and stable labor relations; Sun Zhaoyang and Meng Quan [4,5] argue that at the current stage in China, efforts to build harmonious labor relations with Chinese characteristics must adhere to the concepts of cooperation, balance of rights, balance of interests, and balance of subject logic. Based on the perspective of enterprise management, Mo Shenghong [21] believes that the essence of a harmonious labor relationship is the harmony of the interests of both parties in the labor relationship and the balance of rights and obligations of both parties. Although Chinese and foreign scholars have carried out labor relations studies based on different perspectives, the basic consensus is that harmonious labor relations are a state in which both parties have reached a dynamic balance. On the basis of a comprehensive analysis of relevant literature, this study focuses on the perspective of enterprise management as the basic direction for the evaluation of enterprise labor relations.

2.2. Research on Factors Affecting Harmonious Labor Relations in Enterprises

This paper analyzes the influencing factors of harmonious labor relations in enterprises and provides ideas for refining the evaluation index of harmonious labor relations in enterprises to a certain extent, thus laying a foundation for constructing the evaluation index system of harmonious labor relations in enterprises. Wang Zhijian and Chen Xinmin [22] quantified and analyzed the influencing factors of harmonious labor relations from three levels of enterprises, executives and employees, and believed that the influencing factors include labor contracts, compensation, welfare and social security, labor safety and health protection, employee development and democratic management, labor disputes and mediation; Li Lin and Guo Zan [23], based on the understanding of decent labor, believed that adequate remuneration, sufficient labor protection and adequate development opportunities directly reflect the degree of harmonious labor relations to a certain extent. Cooperative labor relations, as a prerequisite for harmonious labor relations, is likewise an important topic in studies related to the factors influencing labor relations in enterprises. There are two main scales with strong relevance to the measurement of cooperative labor relations: cooperative labor relations climate mainly measures the degree of cooperation in corporate labor relations, including dimensions such as employers’ attitudes toward the employed and the union, the role of union representatives, and the beliefs and behaviors of the employed as corporate participants, and employee participation climate mainly measures the current state of corporate labor relations, including dimensions such as participation in decision making, information sharing, training and performance pay dimensions [24,25,26]; Xu Zelei et al. [27] found that the most important factors in establishing cooperative labor relations include making rationalized suggestions, the number of grassroots union organizations, arbitration and mediation, basic employee health insurance, the implementation of factory affairs in the unit, the establishment of employee representative assemblies, social insurance, work injury insurance, maternity insurance, and implemented rationalized suggestions.

2.3. Research on the Evaluation Index System of Harmonious Labor Relations in Enterprises

Domestic and foreign scholars have conducted a lot of research on the evaluation index system of harmonious labor relations in enterprises and accumulated rich research results. Katz et al. [28] established an enterprise labor relations evaluation system consisting of six indicators: complaint rate, proportion of employees disciplined, number of contracts, negotiation time, absenteeism rate, and attitudes of both employers and employees. Jeffrey B Arthur [29] selected employee participation in corporate governance, resolution of conflicts between employers and employees, degree of decentralization of decision-making power, wages and structure, and the ratio of the number of employees to supervisors as labor relations evaluation indicators. Robert Buchele and Christiansen [30] evaluated labor relations in a particular country or region by measuring three indicators: average years of service, ratio of supervisors to workers, and maximum and minimum earnings of workers. Joseph Wallace et al. [31] used the working environment, wage standard, employee shareholding ratio, job reorganization, job rotation, temporary worker employment, and pension plan as indicators to evaluate the labor relations of enterprises. Jody Hoffer Gittell et al. [32] focused on the role of workers’ subjective feelings in labor relations and divided four categories of indicators, namely union representation, participation in management, labor-management conflict, and workplace atmosphere, to build a labor relations assessment system. Some domestic scholars focus on the interactions between subjects in labor relations, for example, Sun Bo [33] constructed a dynamic and quantifiable evaluation index system of enterprise labor relations from two aspects of management behavior and management results; Xin Benlu [34] introduced subjective indicators such as employees’ identification with the enterprise and constructs a system that includes enterprise management and management, labor contracts, wages and benefits, labor protection, labor disputes, employee self-development, employee identification, rewards and punishments; Yuan Ling et al. [35] established an enterprise labor relations evaluation index system that includes 5 first-level indicators (employment, training and working environment, labor contract and system construction, wage and compensation distribution, social security, trade unions and labor disputes) and 39 s-level indicators; Zhang Huiyan and Jiang Lan [36] constructed an evaluation system that included four basic labor relationship evaluation indexes (establishment of labor relationship, operation of labor relationship, guarantee of labor relationship, and mediation of labor relationship) and five promotion labor relationship evaluation indexes (enterprise care, enterprise management, employee development, reward and punishment, and administrative cooperation).

2.4. Research Review

In summary, scholars at home and abroad have conducted relevant research on enterprise labor relations from different perspectives, which have provided a lot of support for this paper to build an evaluation index system for harmonious labor relations in enterprises. However, there is still some research space in this field, including the following three aspects to be improved. Firstly, the index selection ignores the principle of humanistic demand. The humanism of equality, reciprocity, and mutual respect is consistent with the basic national conditions of a harmonious socialist society with Chinese characteristics. With humanism as the support of the index system, it should be the main direction of future research to trace the factors affecting enterprise labor relations from the root to form the evaluation index of the enterprise labor relations. Secondly, there is a lack of existing research to construct an evaluation index system based on employees’ perspectives. There is a certain cognitive bias in enterprises to the benefit needs of employees, and the relevant investment in labor relations management of enterprises has not been fully transformed into the expected returns. Therefore, constructing an evaluation index system based on employees’ demand perspective can improve and supplement the parts that are easily ignored in employers’ perspective, which is important for realizing harmonious labor relations in enterprises. Thirdly, traditional statistical methods can not deal with latent variables effectively, but structural equation models can analyze latent variables and explicit variables simultaneously by introducing measurement errors. In addition, the traditional weighting method has the defect of ignoring the correlation between the indicators in the dimension level when determining the weight of the index. The structural equation model is introduced to objectively assign the weight to ensure the scientific structure of the evaluation index and the weight of the evaluation index.

From the perspective of employees, this research constructed an evaluation index system for harmonious labor relations in enterprises based on the ERG theory. Firstly, this paper designs the evaluation index system by combining the literature research method and the expert consultation method. Secondly, SPSS 26.0 software and principal component analysis were used to further determine the evaluation index system structure. Thirdly, based on the structural equation model, AMOS 23.0 software is used to carry out first-order and second-order confirmatory factor analysis for the evaluation index system and objectively assign weights to the evaluation indexes according to standardized factor analysis path coefficients. Finally, the above index system is used to accurately evaluate the current situation of labor relations in actual enterprises using structural tests and index weight.

3. Theoretical Framework and Index Design

Via the combing of labor relations-related literature, from the perspective of employees, the existence needs, relatedness needs, and growth needs in the ERG theory were taken as the basic framework of the index system, and the evaluation index system was initially designed using literature research and expert consultation, and then exploratory factor analysis was conducted on the sample data to further determine the evaluation index structure using SPSS 26.0 software and principal component analysis.

3.1. Preliminary Design of Evaluation Index System

Based on the five design principles of operability, scientificity, comparability, systematicity, and representativeness, from the perspective of employees, the ERG theory is introduced into the enterprise harmonious labor relations situation to construct the evaluation index system of enterprise harmonious labor relations. ERG theory, proposed by Clayton Alderfer, professor of organizational behavior at Yale University, is a new humanistic demand theory formed on the basis of revising Maslow’s hierarchy of needs theory. ERG theory divides people’s needs into three core needs, namely existence needs, relatedness needs, and growth needs. Therefore, this paper takes existence needs, relatedness needs, and growth needs as the basic framework of the evaluation index system. In the harmonious labor relations of enterprises, the existence needs of employees refer to the basic material living conditions obtained by individuals via work to ensure life and safety, which corresponds to the philosophical category of id; relatedness needs to refer to the friendly relationship maintained by individuals after they integrate into the workgroup, to meet their social needs and external respect needs, which corresponds to the philosophical category of ego; growth needs refer to the individual’s pursuit of self-worth in his or her career and the satisfaction of his or her internal respect needs at work, which corresponds to the philosophical category of superego. Using the combination of “labor relationship evaluation”, “harmonious labor relationship”, “labor relationship harmony degree”, “labor relationship satisfaction”, and related derivatives as search terms, a large number of documents were consulted, and the evaluation indicators were preliminarily sorted out and summarized from the relevant literature [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48], dimension-level indicators and index-level indicators are determined and divided into corresponding levels. The existence needs of employees generally include stable remuneration and benefits and good working conditions, which are embodied in seven indicators: wage distribution, insurance allowances, flexible benefits, work intensity, work environment, occupational safety and health, and protection of female workers; relatedness needs include good communication management and harmonious interpersonal relationships, which are embodied in 8 indicators: democratic management, labor contract management, corporate culture, development status, industry prospects, individuals and superiors, individuals and colleagues, and organizational atmosphere; growth needs are generally achieved via individual participation and personal development, which are embodied in 6 indicators: decision-making participation, negotiation channels, labor dispute mediation, education and training, career and work significance. By consulting experts in the field of human resource management, the divided evaluation indicators were screened and adjusted, and protection of female workers, development status, and industry prospects was eliminated; the preliminarily designed evaluation index system for harmonious labor relations in enterprises includes a total of 6 dimension-level indicators and 18 index-level indicators, the specific index division and index source are shown in

Table 1. Evaluation index system of harmonious labor relations in enterprises.

Dimension-Level	Index-Level	Indicator Source
salary and benefits (E1)	wage distribution(X1)	Ma Cuihua et al. [10]; Li Yulong et al. [11]; Li Changjiang et al. [41]; Sun Yu et al. [46]; Zhao Haixia [48]
	insurance allowances(X2)
	flexible benefits(X3)
working conditions (E2)	work intensity(X4)	Wei Shun et al. [37]; Sun Yu et al. [39]; Zhou Li [40]; Li Changjiang et al. [41]; Yu Guilan et al. [47]
	work environment(X5)
	occupational safety and health(X6)
communication management (R1)	democratic management(X7)	Ma Cuihua et al. [10]; Zuo Jing et al. [12]; Wei Shun et al. [37]; Sun Yu et al. [39]; Zhou Li [40]; Zhao Haixia [48]
	labor contract management(X8)
	corporate culture(X9)
interpersonal relationship (R2)	individuals and superiors(X10)	Li Yulong et al. [11]; Sun Yu et al. [39]
	individuals and colleagues(X11)
	organizational atmosphere(X12)
individual participation (G1)	decision-making participation(X13)	Li Yulong et al. [11]; Sun Yu et al. [39]; Yu Guilan et al. [47]
	negotiation channels(X14)
	labor dispute mediation(X15)
personal development (G2)	education and training(X16)	Li Yulong et al. [11]; Wei Shun et al. [37]; Sun Yu et al. [39]; Zhou Li [40]; Li Changjiang et al. [41]; Sun Yu et al. [46]; Yu Guilan et al. [47]
	career(X17)
	work significance(X18)

.

3.2. Determination of the Evaluation Index System Structure

3.2.1. Sample Selection and Data Sources

This paper uses the questionnaire survey method to collect data; the questionnaire content is divided into two parts: the first part is the basic personal information, and the second part is to judge the importance of the evaluation index of harmonious labor relations in enterprises. The questionnaires were distributed and collected online, and 232 questionnaires were finally collected. After screening, 210 valid questionnaires were obtained, with an efficiency rate of 90.52%. The nature of the sample’s enterprises, enterprise size, and years of establishment, as well as the gender, education, marital status, and current monthly income of the sample individuals, were analyzed as shown in

Table 2. Enterprise characteristics of the sample.

Enterprise Characteristics	Number of People	Percentage (%)
nature of the enterprise
state-owned business	75	35.7
sino-foreign joint venture	4	1.9
foreign enterprise	14	6.7
private enterprise	77	36.7
other	40	19.0
scale of the enterprise
50 people and below	48	22.8
51–100 people	22	10.5
101–500 people	43	20.5
501–2000 people	26	12.4
2001 people and above	71	33.8
years of establishment of the enterprise
5 years and below	50	23.8
6–10 years	34	16.2
11–15 years	25	11.9
16 years and above	101	48.1

and

Table 3. Sample individual characteristics.

Individual Characteristics	Number of People	Percentage (%)
gender
male	100	47.6
female	110	52.4
education
high school and below	3	1.4
college	29	13.8
undergraduate	141	67.2
postgraduate	37	17.6
marital status
unmarried	98	46.7
married	112	53.3
monthly income
3000 yuan and below	20	9.5
3001–6000 yuan	51	24.3
6001–9000 yuan	82	39.1
9000–12,000 yuan	38	18.1
12,000 yuan and above	19	9.0

. The valid samples collected in the study showed comprehensiveness and diversity in enterprise characteristics and individual characteristics, which provided stability and reliability of the factor analysis results.

3.2.2. Reliability Analysis

This paper uses SPSS 26.0 to process the sample data, and the reliability values of the overall and each dimension are shown in

Table 4. Cronbach’s alpha coefficient table for the overall and each dimension.

Dimension	Cronbach’s Alpha	Based on the Normalization Term Cronbach’s Alpha	Number of Items
overall	0.921	0.921	18
salary and benefits	0.864	0.864	3
working conditions	0.843	0.843	3
communication management	0.889	0.889	3
interpersonal relationships	0.925	0.925	3
individual participation	0.880	0.881	3
personal development	0.816	0.815	3

. The overall Cronbach’s alpha coefficient of the Enterprise Labor Relations Evaluation Index System questionnaire is 0.921, in which the Cronbach’s alpha coefficients for salary and benefits (E1), working conditions (E2), communication management (R1), interpersonal relationships (R2), individual participation (G1), and personal development (G2) are 0.864, 0.843, 0.889, 0.925, 0.881, and 0.815, respectively, all of which were higher than the research standard of 0.7 and had good reliability.

3.2.3. Validity Analysis

This paper uses SPSS 26.0 to calculate KMO and Bartlett values as the data basis for validity, as shown in

Table 5. KMO and Bartlett’s spherical test table.

KMO Measures Sampling Suitability		0.872
Bartlett’s test of sphericity	approximate chi-square	2459.352
	degrees of freedom	153
	salience	0.000

. The KMO detection coefficient is 0.872, greater than 0.6, while the p-value of the spherical Bartlett test is 0.000, less than 0.001. The validity test is passed, and the measurement data effect is satisfactory.

3.2.4. Exploratory Factor Analysis

After six iterations, a total of 6 common factors with eigenvalues greater than 1 are finally extracted, and the cumulative contribution rate of common factors reaches 80.34%. In social science research, the cumulative contribution rate of common factors reaches more than 60%, indicating that common factors can effectively explain all questionnaire variables [49,50]. Considering that the number of selected indicators is large, the coverage is wide, and the selected samples are not only sufficient in number but also have large differences in sample characteristics, it can be considered that the contribution rate of 80.34% has met the standard.

Using SPSS 26.0 to conduct exploratory factor analysis on sample data, using the principal component analysis method, and orthogonally rotating the load factor using the maximum variance method so that each index-level index belongs to a clear common factor, thus forming a reasonable index system dimension structure. The rotated factor loading results are shown in

Table 6. Factor loading results in harmonious labor relations in enterprises.

Indicator Coding	Factor Loadings
	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6
X1				0.807
X2				0.802
X3				0.822
X4					0.766
X5					0.828
X6					0.807
X7		0.832
X8		0.872
X9		0.854
X10	0.820
X11	0.816
X12	0.872
X13			0.795
X14			0.817
X15			0.838
X16						0.823
X17						0.814
X18						0.736

; for the convenience of observation, the values with loading values below 0.5 are deleted from the table.

It can be seen from Table 6 that each indicator belongs to the corresponding dimension, and the load value in another dimension is lower than 0.5. The specific indicators contained in each factor are as follows: factor 1 includes individual and superiors (X10), individual and colleagues (X11), organizational atmosphere (X12); factor 2 includes democratic management (X7), labor contract management (X8), and corporate culture (X9); factor 3 includes decision-making participation (X13), negotiation channels (X14), and labor dispute mediation (X15); factor 4 includes wage distribution (X1), insurance allowances (X2), flexible benefits (X3); factor 5 includes work intensity (X4), work environment (X5), occupational safety and health (X6); factor 6 includes education and training (X16), career (X17), work significance (X18). The extracted six common factors can fully reflect the original data information, which basically shows the distinction and aggregation of the evaluation index system structure. The analysis results show that the division of each index layer index is consistent with the index system designed in Section 3.1, which shows that the evaluation index system of harmonious labor relations in enterprises designed in this paper has structural validity on the whole.

4. Methodology and Hypothetical Model

4.1. Methodology

4.1.1. Structural Equation Model

Structural Equation Model (SEM), as one of the commonly used statistical methods in social sciences, is an important tool for multivariate data analysis. In social science, variables that are difficult to measure directly and accurately are called latent variables, including endogenous latent variables and exogenous latent variables. Traditional statistical methods cannot effectively deal with latent variables, while the structural equation model can analyze latent variables and their explicit variables simultaneously by introducing measurement errors. In the evaluation index system of harmonious labor relations of enterprises constructed in this paper, 18 indexes at the index level, such as wage distribution (X1), insurance allowances (X2), and flexible benefits (X3), etc., are obvious variables. Salary and benefits (E1), working conditions (E2), communication management (R1), interpersonal relationship (R2), individual participation (G1), and personal development (G2) 6 dimension-level indexes are endogenous latent variables, and harmonious labor relations in enterprises are exogenous latent variables. Using the structural equation model, we can analyze the correlation degree between 18 index-level indexes and six dimension-level indexes and the correlation degree between the six dimension-level indexes and harmonious labor relations of enterprises as the target layer. In addition, compared with traditional factor analysis methods, as the structural equation model allows measurement errors of indicators, the standardized factor load of the second-order factor model constructed by using the structural equation model is more accurate and can more truly reflect the internal relationship between indicators. Therefore, the standardized factor load of the second-order factor model of enterprise harmonious labor relations is used to determine the weight of indicators.

This paper uses AMOS 23.0 software to build a structural equation model to carry out confirmatory factor analysis via model fitting, correction, and testing to verify the research hypothesis, to achieve the structural test of the evaluation index system of harmonious labor relations in enterprises, and assign weights to each dimension-level indicator and indicator-level indicator. If the fitting index of the initially constructed model does not meet the standard, the established model needs to be revised until the value of the fitting index reaches the standard. When the value of the fitting index reaches a certain standard, it is considered that a good model fitting effect is obtained, indicating that the constructed hypothetical model conforms to the actual model to the greatest extent. Commonly used fit indices include chi-square degree of freedom ratio (${\chi }^{{}_2}/df$), root mean square error approximately (RMSEA), incremental fit index (IFI), normative fit index (NFI), comparative fit index (CFI), etc. Among them, the ratio of the chi-square degree of freedom ratio (${\chi }^{{}_2}/df$) is preferably between 1 and 3; the closer the comparative fit index (CFI) is to 1, the better, and the smaller the root mean square error approximately (RMSEA), the better.

4.1.2. Basic Structure of Structural Equation Model

(1)

Basic Structure of First-order Confirmatory Factor Analysis

First-order confirmatory factor analysis is used to test whether the relationship between the first-order factors and their observed variables is consistent with the pre-set theoretical relationship and can be used to analyze the relationship between the first-order factors. The model of the first-order confirmatory factor analysis consists of latent variables and their observed variables, and their mathematical expressions are as Formulas (1) and (2):

$$x={\mathsf{\Lambda }}_x\xi + \delta$$

(1)

$$y={\mathsf{\Lambda }}_y\eta + \epsilon$$

(2)

where $\xi$ and $\eta$ are latent variables; $x$ is the observed variable of $\xi$; $y$ is the observed variable of $\eta$; $\delta$ is the observed error vector of $x$; $\epsilon$ is the observed error vector of $y$; ${\mathsf{\Lambda }}_x$ is a matrix of factor loadings of $x$ on $\xi$; ${\mathsf{\Lambda }}_y$ is a matrix of factor loadings of $y$ on $\eta$.

(2)

The Basic Structure of Second-order Confirmatory Factor Analysis

Second-order confirmatory factor analysis is used to test whether the relationship between the first-order factor and the second-order factor is consistent with the pre-specified theoretical relationship. The model of the second-order factor analysis can be represented by the general form of the structural model in the structural equation model in terms of mathematical form, and its mathematical expression is as Formula (3):

$$\eta =\mathrm{B}\eta + \mathsf{\Gamma }\xi + \zeta$$

(3)

Among them, $\mathrm{B}$ is the coefficient matrix describing the interaction between endogenous latent variables $\eta$; $\mathsf{\Gamma }$ is the coefficient matrix describing the influence of exogenous latent variables $\xi$ on endogenous latent variables $\eta$; $\xi$ is the structural residual value.

4.2. Evaluation Index System Hypothesis Model

Based on the theoretical framework in “Section 3.1” and the results of exploratory factor analysis in “Section 3.2.4”, the hypothesis model of the enterprise harmonious labor relations evaluation index system is constructed, as shown in Figure 1. The constructed hypothesis model is a second-order factor model, also known as a higher-order factor model, in which six dimension level indicators of salary and benefits (E1), working conditions (E2), communication management (R1), interpersonal relationship (R2), individual participation (G1) and personal development (G2) are taken as first-order factors and enterprise harmonious labor relations are taken as second-order factors of six dimension level indicators, the six first-order factors are reflected using several observed variables. Specifically, the first-order factor salary and benefits (E1) has three observed variables: wage distribution (X1), insurance benefit (X2), and flexible benefits (X3); The first-order factor working conditions (E2) had three observed variables: work intensity (X4), work environment (X5) and occupational safety and health (X6); The first-order factor communication management (R1) had three observed variables: democratic management (X7), labor contract management (X8) and corporate culture (X9); The first-order factor interpersonal relationship (R2) had three observed variables: individuals and superiors (X10), individuals and colleagues (X11), and organizational atmosphere (X12); The first-order factor individual participation (G1) had three observed variables: decision-making participation (X13), negotiation channels (X14) and labor dispute mediation (X15); The first-order factor personal development (G2) has three observed variables: education and training (X16), career (X17), and work significance (X18). e1–e18 represents the measurement error of the observed variable X1–X18, and e19–e24 represents the measurement error of each first-order factor.

5. Results and Discussion

5.1. Confirmatory Factor Analysis

Using AMOS 23.0 software to build structural equation model and carry out confirmatory factor analysis. The confirmatory factor model constructed in this paper is a second-order factor model. The first-order confirmatory factor analysis is to analyze the degree of correlation between the six dimension-level indicators that reflect the harmonious labor relationship of enterprises from the perspective of employees; the second-order confirmatory factor analysis is to analyze the degree of correlation between the six dimension-level indicators and the harmonious labor relationship of enterprises.

5.1.1. First-Order Factor Analysis

Taking 18 index-level indicators of wage distribution (X1), insurance allowances (X2), flexible benefits (X3), etc. as observed variables, salary and benefits (E1), working conditions (E2), communication management (R1), interpersonal relationship (R2), individual participation (G1) and personal development (G2) six dimension-level indicators are used as first-order factors to construct a first-order initial confirmatory factor analysis model of harmonious labor relations in enterprises, as shown in Figure 2.

After sample selection and data processing, 210 valid sample data were substituted into the initial first-order confirmatory factor analysis model, and AMOS 23.0 was used to fit the model. The fitting results are shown in

Table 7. Fit test table for first-order confirmatory factor analysis.

Fit Index	${\mathit{\chi }}^{{}_2}/\mathit{d}\mathit{f}$	GFI	AGFI	IFI	TLI	CFI	RMSEA
reference value	1~3	>0.9	>0.9	>0.9	>0.9	>0.9	<0.1
first-order initial model	1.413	0.924	0.892	0.980	0.974	0.979	0.044
compliance	up to standard	up to standard	not up to standard	up to standard	up to standard	up to standard	up to standard
first-order modified model	1.278	0.931	0.901	0.986	0.982	0.986	0.036
compliance	up to standard	up to standard	up to standard	up to standard	up to standard	up to standard	up to standard

. The fitting index of the first-order initial model is: the value of ${\chi }^{{}_2}/df$ is 1.413, and the value is between the reference values 1 and 3, which meets the requirements; the values of GFI, IFI, TLI, and CFI are 0.924, 0.980, 0.974, and 0.979, respectively, which are close to 1 and larger than the reference value of 0.9, meeting the requirements; the value of RMSEA is 0.044, which is less than the reference value of 0.1 and meets the requirements; the value of AGFI is 0.892, which is close to 1, but less than the reference value of 0.9, which does not meet the requirements, so the first-order initial model needs to be corrected. In general, the fitting effect of the model can be improved by deleting observed variables that do not meet the requirements, or adding paths between error terms. On the basis of referring to the correction index obtained from the operation of AMOS 23.0 software and considering the practical significance, we choose to establish a path between the error terms e2 and e5 to obtain the first-order correction model. The AMOS 23.0 software was used to fit the data of the first-order correction model, and the fitting results are shown in Table 7. The fitting index of the first-order modified model is: the value of ${\chi }^{{}_2}/df$ is 1.278, the value of GFI is 0.931, the value of AGFI is 0.901, the value of IFI is 0.986, the value of TLI is 0.982, the value of CFI is 0.986, and the value of RMSEA is 0.036, compared with the reference value, it can be seen that the fitting indices of the first-order correction model all achieve the ideal effect. The standardized path coefficients and standardized path diagrams of the first-order modified model are shown in

Table 8. Normalized path coefficients for first-order modified model.

			Estimate
working conditions	<-->	communication management	0.361
working conditions	<-->	interpersonal relationship	0.526
working conditions	<-->	individual participation	0.535
working conditions	<-->	salary and benefits	0.553
working conditions	<-->	personal development	0.541
communication management	<-->	interpersonal relationship	0.496
communication management	<-->	individual participation	0.471
communication management	<-->	salary and benefits	0.496
communication management	<-->	personal development	0.404
interpersonal relationship	<-->	individual participation	0.598
interpersonal relationship	<-->	salary and benefits	0.500
interpersonal relationship	<-->	personal development	0.544
individual participation	<-->	salary and benefits	0.516
individual participation	<-->	personal development	0.476
salary and benefits	<-->	personal development	0.496
X1	<---	salary and benefits	0.862
X2	<---	salary and benefits	0.825
X3	<---	salary and benefits	0.791
X4	<---	working conditions	0.773
X5	<---	working conditions	0.817
X6	<---	working conditions	0.819
X7	<---	communication management	0.895
X8	<---	communication management	0.847
X9	<---	communication management	0.815
X10	<---	interpersonal relationship	0.915
X11	<---	interpersonal relationship	0.864
X12	<---	interpersonal relationship	0.933
X13	<---	individual participation	0.865
X14	<---	individual participation	0.855
X15	<---	individual participation	0.809
X16	<---	personal development	0.832
X17	<---	personal development	0.800
X18	<---	personal development	0.692
e5	<-->	e2	0.388

and Figure 3.

5.1.2. Second-Order Factor Analysis

Salary and benefits (E1), working conditions (E2), communication management (R1), interpersonal relationship (R2), individual participation (G1), and personal development (G2) are six dimension-level indicators as endogenous latent variables, which together reflect a higher-order exogenous latent variable, namely the harmonious labor relationship of the enterprise, needs to continue the second-order confirmatory factor analysis. Taking salary and benefits (E1), working conditions (E2), communication management (R1), interpersonal relationship (R2), individual participation (G1), and personal development (G2) as the first-order factors, the harmonious labor relations of enterprises As a second-order factor, a second-order initial confirmatory factor analysis model is constructed. Before running the model, in order to ensure the smooth operation of the model, the path coefficient between the harmonious labor relationship and salary and benefits (E1) of the enterprise is limited to 1, as shown in Figure 4.

The methods and ideas used in second-order factor analysis are the same as those of first-order factor analysis. AMOS 23.0 is used to fit the data of the second-order initial model, and the fitting results are shown in

Table 9. Fit test table for second-order confirmatory factor analysis.

Fit Index	${\mathit{\chi }}^{{}_2}/\mathit{d}\mathit{f}$	GFI	AGFI	IFI	TLI	CFI	RMSEA
reference value	1~3	>0.9	>0.9	>0.9	>0.9	>0.9	<0.1
second-order initial model	1.401	0.920	0.894	0.979	0.974	0.978	0.044
compliance	up to standard	up to standard	not up to standard	up to standard	up to standard	up to standard	up to standard
second-order modified model	1.269	0.927	0.903	0.986	0.983	0.986	0.036
compliance	up to standard	up to standard	up to standard	up to standard	up to standard	up to standard	up to standard

. The fitting index of the second-order initial model is: the value of The fitting index of the second-order initial model is: the value of A is 1.401, which is between the reference values 1 and 3, which meets the requirements; the values of GFI, IFI, TLI, and CFI are 0.920, 0.979, 0.974 and 0.978, respectively, which are close to 1, and is greater than the reference value of 0.9, which meets the requirements; is 1.401, which is between the reference values 1 and 3, which meets the requirements; the values of GFI, IFI, TLI, and CFI are 0.920, 0.979, 0.974 and 0.978, respectively, which are close to 1, and is greater than the reference value of 0.9, which meets the requirements; the value of RMSEA is 0.044, which is less than the reference value of 0.1, which meets the requirements; the value of AGFI is 0.894, which is close to 1, but less than the reference value of 0.9, which does not meet the requirements, so the second-order initial model needs to be revised. On the basis of referring to the correction index obtained by the operation of AMOS 23.0 software and considering the practical significance, choose to establish a path between the error term e2 and e5 to obtain a second-order correction model. AMOS 23.0 is used to fit the data of the second-order correction model, and the fitting results are shown in Table 9. The fitting index of the second-order modified model is: the value of ${\chi }^{{}_2}/df$ is 1.269, the value of GFI is 0.927, the value of AGFI is 0.903, the value of IFI is 0.986, the value of TLI is 0.983, the value of CFI is 0.986, and the value of RMSEA is 0.036, compared with the reference value, it can be seen that the fitting indices of the second-order correction model all achieve the ideal effect. The normalized path coefficients and normalized path graphs of the second-order modified model are shown in

Table 10. Normalized path coefficients for second-order modified models.

			Estimate
salary and benefits	<---	harmonious labor relations in the enterprise	0.715
working conditions	<---	harmonious labor relations in the enterprise	0.715
communication management	<---	harmonious labor relations in the enterprise	0.618
interpersonal relationship	<---	harmonious labor relations in the enterprise	0.764
individual participation	<---	harmonious labor relations in the enterprise	0.745
personal development	<---	harmonious labor relations in the enterprise	0.694
X1	<---	salary and benefits	0.863
X2	<---	salary and benefits	0.824
X3	<---	salary and benefits	0.789
X4	<---	working conditions	0.770
X5	<---	working conditions	0.814
X6	<---	working conditions	0.823
X7	<---	communication management	0.893
X8	<---	communication management	0.848
X9	<---	communication management	0.817
X10	<---	interpersonal relationship	0.913
X11	<---	interpersonal relationship	0.865
X12	<---	interpersonal relationship	0.933
X13	<---	individual participation	0.866
X14	<---	individual participation	0.853
X15	<---	individual participation	0.811
X16	<---	personal development	0.830
X17	<---	personal development	0.800
X18	<---	personal development	0.696
e5	<-->	e2	0.393

and Figure 5.

5.2. Determination of Indicator Weights

Compared with traditional factor analysis methods, as structural equation models allow measurement errors of indicators, the standardized factor loads of the second-order factor model are more accurate and can more truly reflect the internal relationship between indicators [51,52]. Therefore, this paper uses the standardized factor loads in the second-order factor model of enterprise harmonious labor relations to determine the index weights. Firstly, six first-order factors, including salary and benefits (E1), are normalized to the standardized path coefficients of harmonious labor relations of enterprises, and the weights of indicators of each dimension level are obtained. Secondly, 18 observed variables, including wage distribution (X1), are normalized to the normalized path coefficients of first-order factors, and then the weights of indicators in each index level are obtained. Finally, the relative weights of each index are sorted out and calculated in a summary to obtain the comprehensive weight of the entire evaluation index system.

As can be seen from Figure 5, the standardized path coefficient of the first-order factor salary and benefits (E1) to the harmonious labor relationship of the enterprise is 0.72, then the weight of salary and benefits (E1) is 0.72/(0.72 + 0.71 + 0.62 + 0.76 + 0.75 + 0.69) = 0.170, in the same way, the weights of the remaining five dimension-level indicators can be calculated. The standardized path coefficient of wage distribution (X1) to salary and benefits (E1) is 0.86, then the weight of wage distribution (X1) is 0.86/(0.86 + 0.82 + 0.79) = 0.348, and the comprehensive weight of wage distribution (X1) is 0.170 × 0.348 = 0.059, in the same way, the weights and comprehensive weights of the remaining 17 indicators of the index-level can be calculated. The details are shown in

Table 11. Comprehensive weight of evaluation indicators for harmonious labor relations in enterprises.

Dimension-Level	Weights	Index-Level	Weights	Comprehensive Weight
salary and benefits (E1)	0.170	wage distribution (X1)	0.348	0.059
		insurance allowances (X2)	0.332	0.057
		flexible benefits (X3)	0.320	0.054
working conditions (E2)	0.167	work intensity (X4)	0.321	0.054
		work environment (X5)	0.337	0.056
		occupational safety and health (X6)	0.342	0.057
communication management (R1)	0.146	democratic management (X7)	0.348	0.050
		labor contract management (X8)	0.332	0.048
		corporate culture (X9)	0.320	0.047
interpersonal relationship (R2)	0.179	individuals and superiors (X10)	0.337	0.060
		individuals and colleagues (X11)	0.319	0.057
		organizational atmosphere (X12)	0.344	0.062
individual participation (G1)	0.176	decision-making participation (X13)	0.344	0.061
		negotiation channels (X14)	0.336	0.059
		labor dispute mediation (X15)	0.320	0.056
personal development (G2)	0.162	education and training (X16)	0.356	0.057
		Career (X17)	0.344	0.056
		work significance (X18)	0.300	0.049

.

5.3. Analysis of Application of the Evaluation Index System

In order to reflect the practical significance of the evaluation index system of harmonious labor relations in enterprises, the above-mentioned index system, which was tested by structure and assigned index weights, was applied to evaluate the harmonious degree of labor relations in actual enterprises. The survey is carried out in the form of online questionnaire distribution and collection. The 18 index-level indicators in the system were set up as questions that fit the real working situation, and the evaluation indicators were measured using the Likert scale and divided into five measurement levels: completely disagree, relatively disagree, neutral opinion, relatively agree and completely agree, with the values of 1, 2, 3, 4, and 5 in order. The higher the score, the higher the evaluation of this indicator, and the higher the degree of harmonious labor relations of the enterprise.

According to the evaluation index system, an evaluation model of the harmonious degree of labor relations of enterprises is constructed: $\mathrm{A}={\displaystyle \sum _{i=1}^n{W}_i{B}_i}$ (n = 14), which $W_i$ represents the weight of each index and $B_i$ represents the average score value of the current index. There are five value ranges for the evaluation score of harmonious labor relations in an enterprise: [0,1), [1,2), [2,3), [3,4), [4,5), respectively representing the meaning of enterprise labor relations are discordant, initially discordant, generally harmonious, relatively harmonious, and harmonious. Generally speaking, a score of 3 is regarded as an early warning line for discordant labor relations in an enterprise [48].

Considering the differences in the harmonious degree of labor relations between different types of enterprises, private enterprises, and state-owned enterprises are selected as the analysis objects. N (a private enterprise) and Z (a state-owned enterprise), in the process of long-term operation and development, attach great importance to the construction of harmonious labor relations and sustainable development of enterprises, which is representative to a certain extent. A total of 100 valid questionnaires were obtained from company N and 110 valid questionnaires from company Z after collation. The statistical calculation obtained the index level indicator scores, dimension level indicator scores, and total scores of N and Z companies, as shown in

Table 12. Index scores of enterprise labor relations index level.

Indicator Coding	X1	X2	X3	X4	X5	X6	X7	X8	X9
N company average	3.55	4.20	4.05	3.66	3.65	4.00	3.81	4.2	3.84
Z company average	3.75	4.23	3.97	3.62	3.85	3.87	3.72	4.2	3.85
Indicator coding	X10	X11	X12	X13	X14	X15	X16	X17	X18
N company average	3.97	4.05	3.85	3.82	3.73	3.63	3.71	3.62	3.61
Z company average	3.92	3.96	3.86	3.95	3.77	3.73	3.79	3.55	3.70

and

Table 13. Dimension layer indicator score and total score of enterprise labor relations.

Indicator Coding	E1	E2	R1	R2	G1	G2	A
N company average	3.93	3.77	3.95	3.95	3.73	3.65	3.83
Z company average	3.98	3.78	3.92	3.91	3.81	3.68	3.84

.

According to Table 12 and Table 13, from the perspective of the indicator layer, N company’s wage distribution (X1), work intensity (X4), work environment (X5), labor dispute mediation (X15), career (X17), work significance (X18) scores are all lower than 3.70, and the relative satisfaction is low; the work intensity (X4) and career (X17) scores of Z company are lower than 3.70, and the relative satisfaction is low. From the dimension level, the personal development (G2) score of N company is 3.65, and the relative satisfaction is low; the personal development (G2) score of Z company is 3.68, and the relative satisfaction is low. On the whole, the evaluation scores of the harmonious degree of labor relations between companies N and Z are, respectively ${\mathrm{A}}_{\mathrm{N}}$ = 3.83, ${\mathrm{A}}_{\mathrm{Z}}$ = 3.84, respectively, which belong to [3, 4), indicating that the labor relations between companies N and Z are relatively harmonious, but there is still upside potential.

6. Conclusions and Prospect

6.1. Conclusions

This paper quantifies the concept of harmonious labor relations in enterprises, carries out an empirical analysis of the evaluation of harmonious labor relations in enterprises based on the structural equation model, and finally draws the following research conclusions.

(1)

On the basis of clarifying the relevant concepts and theories of labor relations, combined with the actual situation of harmonious labor relations in Chinese enterprises, this paper took existence needs, relatedness needs, and growth needs of the ERG theory as the basic framework of the evaluation index system, adopted the literature research method and expert consultation method to initially design evaluation index system, and used SPSS 26.0 software and principal component analysis to further determine the evaluation index system structure. Based on the structural equation model, AMOS 23.0 software is used to conduct first-order and second-order confirmatory factor analysis and objective weighting of the evaluation index system. Ultimately, this paper constructed an evaluation index system of harmonious labor relations in enterprises, including six dimension-level indicators and 18 index-level indicators of salary and benefits, working conditions, communication management, interpersonal relationships, individual participation, and personal development.

(2)

Using the index system that has passed the structural confirmation and objective empowerment, the harmonious degree of labor relations in actual enterprises was evaluated. The data of two companies, N and Z, were collected in the form of questionnaires, and the scores of the two companies’ index-level indicators, dimension-level indicators, and overall labor relations were statistically calculated, which were basically in line with the actual situation of enterprise labor relations, indicating that the evaluation index system of harmonious labor relations in enterprises is feasible.

(3)

Compared with previous studies, this study is based on the employee perspective and combines ERG theory to build an evaluation index system, providing a new perspective for the integration of enterprise harmonious labor relations and ERG theory. Using the research of enterprise harmonious labor relations evaluation index system, the harmonious degree of actual enterprise labor relations is accurately evaluated, which provides theoretical guidance for Chinese enterprises to build harmonious labor relations and guarantee the sustainable development of enterprises, and riches the related research of labor relations evaluation in China to a certain extent.

6.2. Limitations and Future Research

The following research deficiencies appear in the research process of this paper.

(1)

Subject to time and manpower constraints, the sample data are still limited, although the total amount of sample data from the questionnaire survey meets the requirements for statistical analysis. The accuracy and applicability of the evaluation index system can be further improved by increasing the sample size of more regions and industries;

(2)

This paper studies the overall enterprise. In order to make a fair, reasonable, and effective evaluation of the harmonious degree of labor relations in different regions, industries, or enterprises of different natures and to achieve the subdivision of research objects, it is necessary to improve and adjust the evaluation index system according to the characteristics of the research objectives, so that its practical application value is further improved.

The research on the evaluation of harmonious labor relations in enterprises can be further carried out in the future from the following two aspects.

Firstly, the change of research perspective. Based on the micro perspective, this paper conducts research on the evaluation of harmonious labor relations in enterprises and constructs an evaluation index system for harmonious labor relations in enterprises from the perspective of employees. In fact, from a macro perspective, it is clear that the subjects involved in enterprise labor relations include the government, other enterprises, the social environment, etc. In addition to paying attention to the relationship between enterprises and employees, the research can be carried out from the point of intervention between enterprises and the government and between enterprises and other enterprises.

Secondly, the change of the research object. Under the background of the sharing economy, platform-based enterprises continue to rise and gradually occupy the dominant position in the business format. Different from the traditional employment model, in the platform-based enterprise employment model, the employment boundaries between workers and organizations are gradually blurred. Fuzzy labor relations are separated from the boundaries of policies and systems, often resulting in the lack of protection of labor rights and social security, and the labor relations of enterprises are prone to disputes and conflicts. Taking the fuzzy labor relationship as the research object, the evaluation of labor relations harmony will become one of the future research directions.