Platform-Based Human Resource Management Practices of the Digital Age: Scale Development and Validation

Abstract

The transformation of organizational platformization provides a technological path and collaborative framework for sustainable development. In this context, platform-based human resource management (HRM) has attracted a lot of attention in academia and the industry, but there is a lack of in-depth research on what dimensions are included in the practice of platform-based HRM and how to measure it. Firstly, this study adopts a theory-based approach to decompose platform-based HRM practices into six functional dimensions, namely “adaptive employee recruitment”, “autonomous job design”, “empowering employee development”, “self-managed compensation management”, “team-based performance management” and “facilitating development planning”. Secondly, based on the scale development procedure, a measurement scale for platform-based HRM practices containing 22 items was developed and passed the reliability test. Finally, the paper conducted a predictive test of the scale with passion for harmonious work as the distal predictor variable and sense of self-determination as the proximal predictor variable, which confirmed the scale’s good predictability. This paper provides a quantifiable tool for related research on HRM in platform-based organizations and offers theoretical guidance and a reference model for building HRM empowerment systems in platform-based enterprises. At the same time, it also provides ideas and references for enterprises to practice platform-based human resources and achieve sustainable development.

1. Introduction

In the digital era, the internal and external environments of organizations have become increasingly complex, ambiguous, volatile, and uncertain. Platform-based organizations have emerged as a new organizational form [1], with platform-based operations within organizations gradually becoming a new model for Organizational Sustainable Development [2]. The change in organizational form will inevitably bring about the change in HRM mode [3]. Within platform-based organizations, the relationship between employees and organizations is no longer “surviving by depending on employment” [4] but has evolved into “equal cooperation”. Therefore, traditional HRM cannot fully be “compatible” with platform organizations [5].

Although supportive HRM, high-engagement HRM, and employee-oriented HRM have recognized the growth needs of employees [6], they do not regard employees as “partners” of the organization, and employees are still “dependent on the organization for survival” in traditional organizations. The purpose of HRM is to cultivate a positive relationship between employees and the organization, and to enhance employees’ sense of identification with the organization [7]. Under this concept, organizations treat employees as a link or part of value creation, which fails to satisfy employees’ needs for self-achievement and also fails to take advantage of employees’ ability to create value as independent individuals in the information age [8]. In the digital era, in order to promote the sustainable development of platform-based organizations, there is an urgent need to construct a set of HRM practices systems that are compatible with them.

In the digital age, organizational management places greater emphasis on the subjective position of employees and the symbiotic relationship between organizations and employees. Some enterprises have begun to explore the platform-based HRM model, such as Haier. In order to adapt to the changes in organizational form of digital age, they constructed a new system of HRM named “convergence and divergence by order”, which has become a collection of successful innovative practices under which the employees are changed from “workers” to “creators” and they share benefits and risks with the enterprise. Reference [9] were among the first to introduce the concept of platform-based HRM. Building on this, Reference [10] suggest that platform-based HRM is a composite construct comprising three main domains: platform empowerment, self-management, and ecosystem governance. Combined with the preliminary explorations of scholars and enterprises, the core concepts of platform HRM can be summarized as follows: Considering employees as equal partners, realizing the organization’s full decentralization of employees, and incorporating empowered employees into HRM, providing employees with the necessary resources and opportunities to become “creators”, and activating employees’ endogenous energy through employee development and reshaping compensation models to achieve self-management of employees. Although scholars have begun to analyze the characteristics of the platform-based HR model [11] and initially developed its measurement scale, there is a lack of analysis of the platform-based HRM practices system from the perspective of the HRM function and developing a matching measurement scale. HRM is a comprehensive system that emphasizes the utilization, development, and cultivation of human resources within an organization through the matching of the functional modules of HRM [12], so it is necessary to study the platform-based HRM practices system from the perspective of management functions.

Based on the above analysis, this paper explores the platform-based Human Resource Management Practices system adapted to era of the digital economy based on the perspective of management functions, summarizes its dimensional composition and the specific practices content contained in each dimension, and develops its measurement scale on this basis. The following study is divided into three parts: firstly, the dimensional structure of platform-based HRM practices is explored using the grounded theory; secondly, the scale of platform-based HRM practices is constructed according to the steps of scale development; and lastly, predictive tests are conducted on platform-based HRM practices.

2. Literature Review

2.1. Literature Review and Commentary

Human resource management practices refer to employee-related tasks performed to provide the number and quality of employees necessary for an organization to achieve its goals [13]. Human resource management practices are typically divided into six areas: recruitment and selection, training and development, performance management, compensation management, career management, and organizational and employee relations [14]. A review of the development of HRM practices reveals that early HRM practices were mostly “control-based” and aimed at reducing labor costs, while Reference [15] proposed a “commitment-based” HRM based on the assumption of human nature in response to the traditional “control-based” HRM practices. Later, Reference [16] further clarified this classification based on the characteristics of strategic HRM practices in firms.

Control-based HRM practices emphasize increasing efficiency by reducing labor costs through strict systems and procedures [17], which inhibit the development of employee autonomy, is not conducive to the sustainable development of organizations. On the other hand, commitment-based HRM practices emphasize the creation of conditions that promote voluntary hard work to achieve organizational goals [18], and therefore can motivate employees to work and mobilize motivation. Following this, scholars have proposed high-involvement HRM [19], high-commitment HRM [20], supportive HRM [21], and developmental HRM [22], which represent a gradual shift from employer-orientated HRM practices to employee-orientated [6], and although these studies began to emphasize the subjective initiative of employees, aiming to enhance the role of employees in organizations, they did not fundamentally depart from the perception of a hierarchical and dependent relationship between organizations and employees, and therefore could not fully realize the self-management of employees.

Platform-based organizations were first proposed by Reference [23]. Platform-based organizations have core characteristics and relative advantages such as bilateral/multilateral markets, network effects, and open boundaries [11]. Reference [24] divides platform enterprises into two types: internal platforms and external platforms. Internal platform enterprises include China’s Haier Group, while external platform enterprises include China’s Alibaba. The platform-based organizations referred to in this paper are internal platform organizations. In recent years, with the increasing prevalence of platform organizations, scholars have attempted to construct new HRM practices systems, but none of them have been able to break away from the perceived limitations of traditional HR practices. For example, the high-involvement HRM practices system constructed by Reference [25] mentions full empowerment of employees but does not clarify the scope of empowerment, which is rather vague, and its reference to reward equity focuses on the equity of remuneration, which can meet employees’ needs for equity but is difficult to act as a motivator. Reference [6] proposed employee-oriented HRM, which emphasizes improving organizational performance by meeting employee needs and affirming employee importance. However, this study only addresses HRM practices broadly without a complete presentation of content. This paper argues that the HRM system should be adjusted in the following three aspects for the transformation of platform organizations: reshaping the organizational structure, reshaping the incentive mechanism, and reshaping the empowerment mechanism.

In this context, Reference [9] introduced the concept of platform-based HRM, which aims to help enterprises explore the construction of self-organization, amoeba teams, dynamic partners and other organizational forms, emphasizing resource sharing and networked management, but does not give a detailed and complete picture of the content of the HRM practices system. Subsequently, Reference [10] constructed a platform HRMS including three sub-systems of empowerment, motivation and authorization based on the AMO model, which is based on the changes in organizational structure, business process, and management logic, and explored the possible proximal and distal outcomes of the system. Unfortunately, the authors did not further explore the composition of each dimension in detail and lacked scientific measurement tools. To compensate for this limitation, Reference [10] followed the rooted theory procedure and summarized platform-based HRM as a construct consisting of the main categories of “platform empowerment”, “self-management”, “eco-governance”, and “ecological governance” as a conceptual model composed of the main categories. However, the model is based on a dimensional analysis of the intrinsic mechanism of human resource management, which has strong theoretical value, while the guidance for practice is relatively weak, and fails to capture the details of platform-based organizations in a complete way. In conclusion, although the existing HRM models or practices have paid attention to creating conditions for employees to voluntarily work towards the achievement of their goals, they are limited to organizational performance and neglect the transformation of organization-employee relationships, especially failing to focus on empowering employees “intrapreneurship and stimulating employees” internal drive as a key focus of HRM, and failing to consider employees’ self-management as the ultimate goal [26]. High-commitment HRM, high-involvement HRM, supportive HRM and developmental HRM in traditional organizations, although partially related to employee “empowerment”, aiming to meet employee needs, improve employee attitudes and behaviors, and thus enhance organizational performance. However, it fails to regard employees as “alliance and reciprocal partners” of the organization, and employees are still “dependent on the organization for survival”, and there is little content advocating internal entrepreneurship among employees. With the platform-based organizational trend deepening, some scholars have begun exploring platform-based HRM systems. For example, Reference [10] examined the dimensions of platform-based HRM practices within the AMO (Ability-Motivation-Opportunity) framework and developed a measurement scale. However, since this analysis is based on the AMO framework rather than the perspective of HRM functional modules, its practical guidance remains limited. In practice, human resource management systems are usually set up with six functional modules: human resource planning, recruitment and allocation, training and development, performance management, compensation and benefits, and employee relationship management [26], so this paper intends to analyze the practice of platform-based human resource management from the perspective of management functions.

2.2. Theoretical Foundation and Conceptual Proposition

Traditional human resource management is confined to the inherent idea of employing and being employed, which causes the internal relationships within the enterprise to lean toward competition than cooperation, thereby diverting attention away from the user; when platform-based organizations abandon the traditional concept of employing and being employed and regards employees as the “dynamic partners” of the organization, the competing relationship will be weakened and the organization will focus its attention on creating value for the employees, which will bring a sustainable internal motivation for the enterprise. According to the reciprocity principle in social exchange theory [27], when a platform-based organization empowers its employees, provides them with entrepreneurial opportunities and related resources, and treats them as equal partners, they gain the organization’s “support” and “respect”. Driven by a “reward” mentality, the employees will actively manage their personal careers, engage in self-management, and promote the sustainable development of the organization in the process of realizing their own value. As platform-based HRM practices emphasize respect for employees’ humanity [9,26], this paper compares them with the two typical human resource management models (control-based HRM and commitment-based HRM) proposed by Reference [15] from the perspective of human nature assumptions, as well as with “developmental HRM”, which shares similarities with platform-based human resource management (as shown in

Table 1. Comparison of HRM practice systems.

	Control-Based HRM	Commitment-Based HRM	Developmental HRM	Platform-Based HRM
Employee needs	Material incentive	Material incentives + Motivational needs	Growth, development	Achievement needs + Self-drive + User value
Practice-oriented	Employer-oriented	Employee-oriented	Building a development platform	Employee-driven + Market-oriented
Focus of attention	Production management	Employee relations	Employee development opportunities and aspirations	Platform empowerment
Practical objective	Reducing labor costs	Strengthening employee organizational commitment	Improvement of value creation capacity	Enhancing organizational endogenous motivation

Note: The data in this table have been compiled by the authors.

).

Control-based HRM is a traditional HRM approach [12], which focuses on organizational effectiveness and material incentives but ignores employees’ needs and aims to reduce labor costs, which becomes more and more ineffective with the increase the in employees’ own needs. On the other hand, commitment-based HRM is relative to control-based HRM, as it begins to pay attention to employees’ needs, hoping to make them actively work, undertake extra work tasks driven by “organizational commitment”, and promote the improvement of organizational performance [12]. This process turns employer-oriented human resource management into employee-oriented, which has been an advocated HRM model in recent years. Compared with these two typical models, platform-based HRM has made a subversive change in the perception of the relationship between enterprises and employees. It views employees as partners and the organization as a platform supporting employee development, positioning the HRM department as a “service role” that empowers employees. In this context, platform-based HRM is no longer limited to “improving organizational performance” but is based on the platform’s integrated resources, which provide employees with equal entrepreneurial opportunities to truly work for themselves enabling them to realize their self-worth while simultaneously promoting the development of the organization.

While developmental HRM focuses on “collaborative development”, platform-based HRM places greater emphasis on “employee entrepreneurship”, driving the long-term development of the organization through employees’ self-management. In particular, platform-based HRM regards employees as partners, which more prominently highlights the core value of employees. It emphasizes binding the interests of employees and the organization through internal entrepreneurship to make them partners. In addition, platform-based human resource management differs fundamentally from the digital human resource management practices [28] and computational human resource management [29] recently proposed by scholars. Digital human resource management practices emphasize data-driven human resource management decisions, while computational human resource management emphasizes data algorithms to assist human resource management decisions. Platform-based human resource management practices emphasize that the organization provides a work platform for employees to achieve mutual growth.

In summary, this paper defines platform-based HRM as a set of HRM practices system designed to regard employees as equal partners and the enterprise as a platform that empowers employees’ development, promote employees’ self-worth realization, and cultivate the organization’s endogenous drive by means of empowerment. The HRM practices start with “platform empowerment” and aim to achieve self-management of employees by stimulating their internal motivation.

3. Methodology and Analysis

Based on the definition of platform-based HRM, separate interviews with weaving staff and managers of the platformization group are conducted below. The relevant contents of HRM and employees’ subjective perception evaluation in typical platforming enterprises are collected. We refer to articles by other scholars to obtain complete and detailed information about organizations and employees. We explore the characteristics of platform-based HRM practices in specific management functions and how they affect employee motivation and behavior, on the basis of which a measurement scale based on the perspective of management functions is developed and its predictive effect is tested. Therefore, this study is divided into three parts.The first part explores the structure of the functional dimensions of platform-based HRM practices, the second part develops a measurement scale for platform-based HRM practices, and the third part tests the predictive effects of platform-based HRM practices. Therefore, the research framework of this paper is shown in Figure 1.

3.1. Conceptual Development

Grounded theory is a bottom-up approach to theory construction, characterized by its rigorous methodology, traceability, and replicability [30]. This method allows for exploration without being constrained by existing theories, making it well-suited for exploratory research in new domains. Accordingly, this section applies grounded theory, following the process of “open coding—axial coding—selective coding” [31], based on theoretical sampling, to clarify the underlying mechanisms in practice. Through this approach, the structural dimensions of platform-based HRM practices are identified, and the internal mechanism of “platform empowerment—self-determination—work motivation” within platform-based HRM is preliminarily established.

3.1.1. Theoretical Sampling and Data Acquisition

Based on the theoretical sampling principle of grounded theory, factors such as human resource management concept, industry classification, and enterprise size were included in the selection criteria. Drawing on past experience [11], the interviewed enterprises were platform organizations with more than 150 employees and had been carrying out human resource management practices for more than one year. After the screening process, 14 platform organizations were finally selected, involving 8 industries such as home appliance manufacturing, gaming, department store retailing, information industry, and garment design.

In line with grounded theory procedures, the data were collected through multiple channels from December 2020 to July 2023 to ensure data triangulation. First, a semi-structured interview method was used to investigate employees and managers of the company, focusing on topics such as platform-based employee empowerment and intrinsic motivation. Sample questions included “What HRM practices does your company implement to enhance employees’ comprehensive abilities?” and “How do you think HRM should handle the relationship between the organization and employees in a platform-based organization, and how is your company approaching this?”. The interviewees were from 14 platform-based companies representing the 8 industries mentioned above. These companies are located in Shandong, Jiangsu, Zhejiang, and Guangdong, among others. We used our alumni and classmate relationships to reach out to the interviewees. The sample consisted of 23 managers and 25 employees, including 27 males and 21 females, aged between 22 and 45, with an average age of 37.6. We went to the frontline of the business to conduct interviews. The employee interviews typically lasted for 25–50 min, while the length of the management interviews ranged from 45 to 90 min, with an average duration of approximately 48 min. Secondly, in order to make the collected interview data reflect the characteristics of HRM practices under the platform organizational change: the non-platform organizations that the focal firms are benchmarked against were identified, and the HRM practices collected from the official websites of the focal firms, public media, and other channels were compared with those of the benchmark firms, then the content of the HRM practices, in line with the platform organizations, was selected. Finally, in order for the information to satisfy completeness and saturation, additional data were gathered from CNKI, following Reference [32], including case studies, articles on organizational structure, HRM, and employee relations for the selected target companies, as well as online reports to supplement and enhance the collected data.

In total, the interviews amounted to 2332 min. After transcription and proofreading, the interview data amounted to 44,015 words, with a total of 46,952 words of cumulative textual data. Following the approach of Reference [11], 75% of the textual data was selected for coding, with the remaining data used to test for theoretical saturation.

3.1.2. Coding Process

First, we carried out open coding. In order to make the initial concepts as relevant as possible to the reality of platformized HRM practices, the open coding in this study used the respondents’ original codes to distill the corresponding concepts as much as possible [31]. In addition to coding platform-based HRM practices, this study also coded how the practices system affects employee motivation and behavior. During the coding process, the initial concepts were first coded independently by three postgraduate students to form initial concepts and eliminate conflicting codes. Then, the three graduate students pooled and discussed similar codes (the three postgraduate students are the last three authors of this paper, and they have given their informed consent to the paper), categorizing and merging them to form 81 concepts and 37 categories (

Table 2. Open coding (example).

Category	Initial Concept	Original Material
High work freedom	Flexible work progress, individual decision on work content	A13: We are not involved within their teamwork, as to what extent we do it and exactly what we do on which day, that is not something we should be concerned about; we just look at the final result. B7: It is up to us to decide what we should do and to what extent we should do it when we carry out our work, certainly within the committed deadlines.
Pursuit of person-job fit	Big data job analysis, comprehensive consideration in job assignment	A2: We will use big data to analyze our staff and try to put everyone in the right position so that we can maximize everyone’s potential. A5: There are a lot of factors to consider when allocating positions, not just what kind of person the company needs as before, but also whether the person is suitable for the position, as well as their own expectations, etc. We will also refer to the results of the big data matching.
Stock Allocation for Outstanding Members	Equity incentives, working for oneself	A9: If a team is doing a very successful job, we can allocate different shares to the team members according to the size of their contribution, so that the employees can really work for themselves. B2: There are a few highly successful teams to whom the company gave shares, allowing them to become working for themselves—the drive is naturally different.
Team Autonomy	Team self-hiring, autonomously deciding salary distribution	A12: The hiring of personnel within each team is an internal matter for their team, which is free to decide who needs to be hired or dismissed, etc. B3: After the completion of the team project, excluding the company’s share, the remaining portion can be freely allocated—we can distribute dividends or proceed to the next step.
……

Note: The data in this table have been compiled by the authors.

). Finally, after completing the open coding process, we invited professors from the field of human resource management (these experts came from Northwestern University, Capital University of Economics and Trade, and Ocean University of China in China), as well as HR supervisors and managers from various enterprises (they came from China’s Haier Group, Aucma Group and Sailun Group, among others) to review the results. Their feedback was incorporated, and iterative modifications were made to ensure the reliability of the findings.

Second, we carried out axial coding. The categories obtained from the open coding were linked together in conjunction with the usual criteria for dividing HRM functions and drawing on the exemplary model of procedural rooting theory [31]. After screening, categorizing and merging, the 39 categories were grouped into 9 main categories, 6 of which belonged to the dimensional structure of HR practices, and were highly compatible with the six functional modules of the HRM practices system; the other three main categories belonged to the results of HRM practices (

Table 3. Selective coding results.

Main Category	Category	Initial Concept
Autonomous job design	Project self-organization	Freedom to determine the pace and content of work while working
	Freedom of office	Considerable freedom in the way work is carried out in the workplace
	Freedom to handle issues	Employees have autonomy to deal with problems at work within the scope of their work
	Autonomy of work time and space	Irregular working hours, employees can decide freely
	Flexible job options	Employees can choose a new job if they find their current position unsuitable
Empowering employee Development	Basic content training	Employees should familiarize themselves with the company’s culture, values, and jobs through micro-learning videos, etc.
	Skills upgrading training	After completing the basic training, employees can choose other micro-lesson videos on the training platform for learning
	Provision of targeted training	Analyzing employees’ potential skills to achieve precisely targeted training
	Focus on the integration of practical training	Enhancement of talent through a combination of training and practice
	Helping to match people with jobs	Use big data to accurately match jobs with employees’ strengths and personalities
Self-managed salary management	Team pay is self-supporting	Employees’ main salary comes from self-employed organizations
	Excess profit sharing	Employee participation in sharing when projects realize value and create excess profits
	Self-financed and self-supporting	Start-up teams are required to self-fund their monthly living expenses for a limited period of time
	Difference-sharing incentives	Sharing rates vary at different points in excess of expected profits
	Pay-for-performance betting	Employees use part of their salary to gamble with the company and can be returned according to a certain multiplier if they meet the performance requirements
	Customer benefits	Share of proceeds is linked to user ratings
Team-based performance management	Self-defined performance levels	Team autonomy in determining team performance levels
	Team rights self-determination	The team can decide on the retention of personnel and the distribution of salaries
	Performance allocation endogenous	Based on team performance, each employee decides their own personal performance
	Risk-return sharing	Team member earnings are linked to overall team benefits
	Performance competition orientation	Reduce emphasis on employee evaluation and focus instead on their practical performance, so that talent is selected based on actual ability rather than credentials alone
Boosting development planning	Basic competence development	New employees acquire job skills through job enlargement and job rotation
	Facing market needs	Employees can leverage company resources to discover user pain points and form teams to propose solutions
	Freedom of direction	Teams can decide where they want to go by agreement with the company once the team is formalized
	Endogenous motivation of employees	Distribute equity to the members of the best teams
Adaptive employee recruitment	Recruitment expanded	Enterprises can use the internet to recruit talent nationally and globally
	Combination of internal and external recommendations	Organizations encourage internal employees and external stakeholders to recommend talent
	Implementation of precision recruitment	Use of big data and other technologies to accurately recruit and job-match talent
	Focus on competence and integrity	Selection of staff is based on the candidate’s personal ability and integrity and nothing else
	Employment sector autonomy	Hand-picked by the hiring department to meet the needs of their team
Psychological needs	Employee autonomy requirements	Flexibility in dealing with work issues within the scope of work
	Employee relations needs	Employees can be recognized and respected by others at work
	Staff development needs	Employees can improve their overall personal competence in the course of their work
Goal integration	Benefit level sharing	Employees are rewarded for their contribution to the organization
	Symbiosis in the development dimension	Organizations and employees can promote and develop each other in the course of their work
	Spiritual co-prosperity	Employees recognize and strive for the organization’s mission and vision
Motivation	Strong interest in work	Employees are motivated and engaged because of their interest in their work
	Great enthusiasm for work	Employees are passionate about the work they are doing
	Taking the initiative to overcome difficulties	Employees will take the initiative to find ways to overcome difficulties at work

Note: The data in this table have been compiled by the authors.

).

Among them, “autonomous job design” refers to the job design within the framework of organizational rules through the authorization mechanism, enabling employees to independently control the work rhythm and decision-making process; “empowering employee development” refers to the human resource management function of cultivating employees’ independent decision-making ability and growth drive through the mechanism of ability activation and management authorization; “self-managed salary management” refers to a model of salary management that matches the salary of employees with the user value created through the introduction of market mechanisms; “team-based performance management” refers to a performance management mode that takes the team as a whole as an appraisal unit through systematic management activities such as target synergy, process monitoring and result evaluation; “boosting development planning” refers to the human resource allocation method that realizes the added human capital value and enterprise efficiency enhancement through systematic talent development channel design and resource allocation; and “adaptive employee recruitment” is a new type of talent selection mode with multi-dimensional matching as the core, emphasizing the dynamic adaptation of candidates to positions, teams, and organizational culture.

Third, we carried out selective coding. Selective coding further linked different categories together, considering the logical relationships between categories, describing behavioral phenomena in a narrative manner, identifying core categories, and constructing a theoretical framework [30]. After selective coding, “autonomous job design, empowering employee development, self-managed salary management, team-based performance management, boosting development planning, and adaptive employee recruitment” were grouped into the core category of “platform-based HRM practices”, while “psychological needs, goal integration, and motivational performance” were grouped into the core category of “employee motivation”. The HRM mechanism of “platform empowerment—endogenous drive—self-management” was summarized on this basis. In this mechanism, the platform HRM practices system took autonomous job design, empowering employee development, self-managed salary management, team-based performance management, boosting development planning, and adaptive employee recruitment as the input of HRM practices, and took the self-determination mechanism that met the psychological needs of employees’ development, autonomy, and relationship, and achieved the integration of organizational and employee goals as the process, so as to realize the output of stimulating employees’ intrinsic motivation, realizing employees’ self-management, and enhancing their work enthusiasm (Figure 2).

Fourth, theory saturation testing. Theory saturation testing can be conducted when no more new categories emerge from the interview data [30]. In this study, the remaining interview data were used to test theory saturation, and the results were compared with the established results. It was found that no new concepts or categories emerged, indicating that the theory was saturated.

Fifth, reliability and validity testing. This study was independently coded by three experienced postgraduate students to reduce personal subjective bias. The results were aggregated and merged, and then professors in human resource management-related fields and supervisors and managers in charge of human resource management in the enterprises were invited to view the results and make modifications to the results, so as to ensure that the results were of good reliability. Before the research began, this study developed a series of screening criteria, such as enterprise size, time of implementation, and HRM practices.

3.2. Scale Development

3.2.1. Compilation of the Initial Scale

After a grounded theoretical study, an initial scale containing 30 measurement entries was initially developed. On this basis, according to the conventional process of scale development [33], volunteers were selected to conduct an open-ended questionnaire survey, in which the definition of platform-based HRM practices were first presented to the subjects and then served as the material for the additional questions based on the subjects’ descriptions. The open-ended survey was conducted in the relevant departments of Haier and TGOOD, and a total of 52 employees were recruited to participate in the interviews, which mainly included “what kind of platform the company should provide to help employees develop and thus improve organizational performance”, “what are the advantages of empowering employees through HRM”, “how to carry out HRM to empower employees”, and so on. Based on the descriptions of the employees participating in the interviews, semantic segmentation, labeling, summarization, deletion, and other processes were used to exclude semantic duplications with the previous 30 entries, and 6 new entries were added.

In addition to the open-ended questionnaire, we continued to search the literature related to platform HRM in CNKI, Web of Science, EBSCO and other databases, extracting concepts similar to the six dimensions of the grounded theory process, such as competence development practices, flexible work design, participatory pay management, and screening relevant items that can be incorporated into the aforementioned 6 dimensions. On the basis of the previously collected entries, 7 new entries were added, resulting in a total of 43 items.

A total of 5 professors who had been engaged in HRM research for many years were invited to evaluate the initial scale developed in this paper to guarantee that the content validity meets the criteria. Firstly, the professors were told about the platform-based HRM practices and the meaning of each dimension to give them an in-depth understanding of the concepts proposed in this study. Then, the professors were asked to judge whether the test items of the scale matched the concepts and whether the presentation was reasonable and free of ambiguity. After comprehensively listening to the professors’ suggestions, the relevant items of the scale were modified.

Subsequently, 16 HR leaders of platform-based companies were invited to evaluate and score the scale using a 5-point rating system. ① Analyze to what extent the test items of the scale can promote employee self-motivation; ② judge whether the test items of the scale are reasonable and practical. Finally, we received evaluation letters from the human resource managers of 11 enterprises. From the results, each test item met the requirements, forming a 43-question pre-test scale, of which the autonomous job design items are Q1–Q7, the empowering employee development items are Q8–Q15, the self-managed salary management items are Q16–Q23, the team-based performance management items are Q24–Q30, and the boosting development planning items are Q31–Q36 and Q31–Q36, and adaptive employee recruitment items are Q37–Q43.

3.2.2. Scale Purification and Exploratory Factor Analysis

First, data collection. The scale was evaluated using a 5-point Likert scale, and the source of the questionnaire was divided into two parts: firstly, the paper offered a questionnaire to participants from platform companies who were pursuing MBAs at universities. They were asked to answer and recover them on the spot. Secondly, an electronic letter with an attached electronic questionnaire was sent to the HR supervisors or managers of platform-based enterprises in various industries through the social connections of the teachers in the school to ask for their help in forwarding the questionnaires to the other HR supervisors or managers of other platform-based enterprises, and then we collected them after 10 days. After counting, 70 paper questionnaires were distributed and 70 were recovered; 256 electronic questionnaires were distributed and 234 were recovered; and 263 valid questionnaires were obtained after screening 304 (questionnaires 70 + 234). The total number of the interviewed companies was 198, mainly distributed in Qingdao, Jinan, Tianjin, and Zhengzhou. Among the respondents, 45.6% were male and 54.4% were female. In terms of age distribution, 35.3% were aged 20–30, 52.8% were aged 30–40, and 11.9% were over 40. Regarding educational background, 59.6% held an associate’s or bachelor’s degree, 28% held a master’s degree, and 12.4% held a doctoral degree. For work experience, 13.5% had less than 1 year, 40.7% had 1–3 years, 23.8% ad 3–7 years, 14.2% ad 7–10 years, and 7.8% had more than 10 years.

Second, the screening of pre-test scale items. The critical ratio value of each item was calculated, and the items Q4, Q5, Q8, Q13, Q15, Q18, Q21, Q23, Q24, Q29, Q31, Q32, Q39, Q42, and Q43, totaling 15 items, which did not reach the level of significance, were deleted. The remaining 28 items were analyzed again for reliability and discrimination. The results showed that the Cronbach’s α of the scale as a whole was 0.889, indicating that the reliability of the scale meets the requirements; the CITC value was used to judge the reliability of the test items, and it was found that the Cronbach`s α value was lower than 0.884 after any of the question items had been removed, which indicated that each question item was reliable, and therefore each question item should be retained. Subsequently, two categories of “high” and “low” scores were set according to the total scores of the 28 question items, and the analysis found that the differences between the “high group” and the “low group” were significant (p < 0.001), indicating that the discriminatory ability of the scale items meets the requirements.

Third, factor loading analysis. Using SPSS 25.0, the sample data were subjected to the KMO test and Bartlett’s test of sphericity. The KMO value was 0.835, and Bartlett’s test was significant at the 0.001 level, indicating that the data were suitable for exploratory factor analysis. Following the approach of Reference [32], principal component analysis with varimax rotation was employed to extract factors with eigenvalues greater than 1. The items were removed if they did not meet the following criteria: ① factor loadings below 0.5 and ② cross-loadings above 0.35. Based on this analysis, 6 items (Q6, Q11, Q20, Q26, Q30, Q34) were eliminated. A subsequent factor analysis on the remaining items produced a six-factor, 22-item structure, as shown in

Table 4. Results of exploratory factor analysis.

Project	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6
AJD1: We have the freedom to decide the pace and content of our work (A)	0.764	0.135	0.191	0.123	0.169	0.077
AJD 2: We can decide how to carry out our work (A)	0.823	0.130	0.162	0.127	0.155	0.135
AJD 3: We can handle issues within our authority independently (A)	0.840	0.133	0.162	0.171	0.152	0.133
AJD 4: The company is like an enabling platform, with more services than constraints (B)	0.794	0.112	0.189	0.121	0.216	0.042
EED1: After completing basic training, employees can select additional micro-courses for learning on the platform (B)	0.206	0.836	0.098	0.132	0.248	0.040
EED2: The company combines training with real projects to develop our skills (B)	0.241	0.807	0.152	0.15	0.133	0.035
EED3: The company emphasizes precise matching between positions and personal strengths or personalities (B)	0.240	0.768	0.124	0.175	0.134	0.040
EED4: The company provides rich training resources and learning opportunities for employees to choose independently (A)	0.275	0.780	0.130	0.160	0.197	0.125
SCM1: Our income is closely linked to the performance of our own business operations (A)	0.199	0.131	0.804	0.116	0.191	0.057
SCM2: The company shares excess profits from specific projects with us (C)	0.261	0.097	0.839	0.138	0.077	0.059
SCM3: Our share of earnings is linked to customer evaluations (C)	0.284	0.189	0.782	0.145	0.141	0.134
SCM4: In terms of compensation management, it feels like we’re working for ourselves (C)	0.181	0.062	0.820	0.096	0.18	0.119
TPM1: Our team can decide on performance levels based on actual conditions	0.142	0.067	0.110	0.832	0.284	0.189
TPM2: The company grants the project team full internal performance management authority	0.210	0.149	0.102	0.855	0.181	0.062
TPM3: The company mainly conducts performance management on a team basis	0.064	0.177	0.127	0.819	0.206	0.205
FDP1: The company enables us to quickly master job skills through job expansion and rotation (B)	0.135	0.075	0.148	0.135	0.861	0.165
FDP2: The company encourages and empowers employees to create internal startups within teams (B)	0.022	0.101	0.09	0.022	0.860	0.129
FDP3: The company provides ample or continuous opportunities for advancement (B)	0.125	0.132	0.005	0.125	0.834	0.110
AER1: Our company encourages internal employees or suppliers to recommend talent (B)	0.197	0.221	0.108	0.255	0.197	0.798
AER2: The company recruits employees without restrictions, focusing on matching skills to roles (C)	0.205	0.098	0.132	0.248	0.205	0.806
AER3: The company has diverse recruitment methods (A)	0.217	0.152	0.15	0.133	0.217	0.822
AER4: The company provides sufficient recruitment authority to the hiring departments (A)	0.206	0.124	0.175	0.134	0.206	0.816

Note: The data in this table have been compiled by the authors.

, with a cumulative variance explanation rate of 74.029%, and factor loadings ranging from 0.761 to 0.861. The dimensions, grounded in the results of the grounded theory analysis, were named as follows: autonomous job design (AJD), empowering employee development (EED), self-managed salary management (SSM), team-based performance management (TPM), boosting development planning (BDP), and adaptive employee recruitment (AER).

When exploring the dimensions of platform-based HRM practices using grounded theory, this paper used prior knowledge of management functions for axial coding, resulting in 6 main categories, while Reference [10] used the AMO framework model for axial coding, resulting in 3 main categories; therefore, there are differences in the dimensions obtained by the two studies. In order to verify whether there is an inherent consistency between the two studies, this paper used the factor number setting function of SPSS to set the 22 question items mentioned above, and the factor after dimensionality reduction was set to 3. Exploratory factor analysis was conducted again, and it was found that the goodness-of-fit obtained for the 3 factors also met the standard requirements (χ²/df = 1.831, RMSEA = 0.022, CFI = 0.956, TLI = 0.955, SRMR = 0.034). As shown in Table 4, the question items corresponding to the three factors are labeled A, B and C. A comparison with Reference [10] indicates that the items across both studies are generally consistent.

3.2.3. Confirmatory Factor Analysis

First, data collection. Exploratory factor analysis can initially delineate the dimensions of platform-based HRM practices but cannot analyze the overall fit of the factors [26]. Therefore, the data were collected again to test the overall fit of the factors through confirmatory factor analysis. Letters and electronic questionnaires were sent to the managers of 105 platform-based enterprises nationwide through both online and offline channels, and a total of 248 questionnaires were collected, and 215 valid questionnaires were obtained after screening. The interviewed group accounted for 43.7% of men and 56.3% of women; 46.1% were 20–30 years old, 48.3% were 30–40 years old, and 15.6% were 40 years old or above; 65.5% had an associate or bachelor’s degree, 21.7% had a master’s degree, and 6.8% had a doctoral degree; 9.4% had worked for less than 1 year, 35.7% for 1–3 years, 23.5% for 3–7 years, 12.6% for 7–10 years, and 18.8% for more than 10 years.

Second, the common method bias test. The Harman single-factor test analysis method was used to conduct exploratory factor analysis on the measurement items of each variable. The results showed that the first factor extracted from the unrotated exploratory factor analysis accounted for 24.478%, which was less than half of the total explained variance. This indicates that there is no serious common method bias issue, and the collected data is representative.

Third, process and result analysis. Confirmatory factor analysis of the 22 test items was conducted using Mplus 7.4 software, followed by the setting of alternative models. (1) A four-factor model: autonomous job design and adaptive employee recruitment were each one factor, while self-managed salary management and team-based performance management were combined into one factor, boosting development planning and empowering employee development were also merged into one factor. (2) A five-factor model: self-managed salary management, team-based performance management, autonomous job design and adaptive employee recruitment were each one factor, and boosting development planning and empowering employee development were merged into one factor. (3) A six-factor model: six factors of self-managed salary management, team-based performance management, autonomous job design, empowering employee development, boosting development planning, and adaptive employee recruitment.

Table 5. Fit indices for different models.

Model	χ2/df	RMSEA	CFI	TLI	GFI	SRMR
Four-factor model	4.958	0.113	0.778	0.748	0.512	0.093
Five-factor model	3.160	0.083	0.881	0.862	0.701	0.068
Six-factor model	1.050	0.013	0.967	0.967	0.829	0.031
Second-order six-factor model	1.044	0.012	0.968	0.967	0.832	0.035

shows that the six-factor model demonstrated the best fit χ²/df = 1.050, RMSEA = 0.013, CFI = 0.997, TLI = 0.997, GFI = 0.829, SRMR = 0.031), and that all fit indices met the requirements. The previous study showed that the six dimensions may form a higher-order variable, and after grouping the six factors into a second-order factor model, each fit indices met the criteria (χ²/df = 1.044, RMSEA = 0.012, CFI = 0.998, TLI = 0.997, GFI = 0.831, SRMR = 0.035). This finding indicates that platform-based HRM practices consist of six first-order factors—self-managed salary management, team-based performance management, autonomous job design, empowering employee development, boosting development planning, and adaptive employee recruitment—comprising a second-order factor structure.

3.2.4. Reliability and Validity Testing

First, reliability testing. Using data collected from the second round, the overall Cronbach’s α and composite reliability (CR) of the scale and its dimensions were analyzed to examine internal consistency. The results showed an overall Cronbach’s α of 0.900 for the scale, with reliability coefficients of 0.862, 0.858, 0.870, 0.863, 0.879, and 0.866 for autonomous job design, empowering employee development, self-managed salary management, team-based performance management, boosting development planning, and adaptive employee recruitment, respectively. Additionally, all dimensions had CR values above the standard 0.8, indicating strong internal consistency and passing the reliability test.

Second, validity testing. This study followed grounded theory steps to thoroughly compare interview data and the literature, and multiple professors in HRM and HR managers from companies were invited to evaluate the questionnaire content. Based on their feedback, the questionnaire items were revised, resulting in a platform-based HRM practice scale with 6 dimensions and 22 items, thus confirming good content validity. The results of the convergent validity analysis (

Table 6. Correlation Analysis and Reliability Analysis of Dimensions.

	V1	V2	V3	V4	V5	V6	AVE	Cronbach’s α	CR
AJD: Autonomous job design	0.785						0.617	0.862	0.865
EED: Empowering employee development	0.350 **	0.778					0.606	0.858	0.860
SSM: Self-managed salary management	0.295 **	0.356 **	0.792				0.628	0.870	0.871
TPM: Team-based performance management	0.329 **	0.364 **	0.382 **	0.824			0.679	0.863	0.864
FDP: Boosting development planning	0.365 **	0.383 **	0.318 **	0.311 **	0.843		0.710	0.879	0.880
AER: Adaptive employee recruitment	0.337 **	0.354 **	0.358 **	0.365 **	0.328	0.791	0.625	0.866	0.870

** p < 0.01 and numbers of boldface in diagonal are square roots of AVE.

) indicated that the AVE values for the six factors—empowered employee development, self-managed salary management, team-based performance management, facilitated development planning, and adapted employee recruitment—ranged from 0.606 to 0.710, all exceeding the 0.5 threshold. Moreover, the second-order confirmatory factor analysis revealed that the loadings of the six first-order factors on the second-order factor ranged between 0.641 and 0.688, with a good model fit, thus confirming convergent validity. Table 6 also shows that the correlation coefficients between any two dimensions were all less than the square root of the AVE, indicating good discriminant validity among the six factors. In summary, the construct validity of the scale is robust.

4. Results

4.1. Research Hypothesis

According to the “platform empowerment–self-determination–work motivation” mechanism model constructed in Figure 2, it can be seen that platform-based human resource management practices not only meet employees’ psychological needs but also promote the integration of organizational and employee goals, thereby stimulating employees’ intrinsic work motivation. Relevant studies have pointed out that the satisfaction of psychological needs and the integration of organizational and employee goals makes employees have a sense of self-determination [34], and at the same time stimulates work motivation and spontaneous behavior [35]. Meanwhile, employee motivation and self-management can be achieved through the harmonious work passion of the employees who “spontaneously love their work from the bottom of their hearts” [36]. Therefore, this paper chooses self-determination as the proximal predictor variable and harmonious work passion as the distal predictor variable to test the mechanism of Figure 2.

Work passion reflects a strong inclination towards work itself, focusing on the cognitive and emotional factors involved when employees are engaged [37]. Harmonious work passion, specifically, is rooted in an intrinsic love for the work rather than external pressures or outcomes; it is a sustainable work motivation [35]. Previous studies have found that platform-based organizational mechanisms can stimulate employees’ work passion; for example, Reference [35] found that job–person fit and resource empowerment in platform organizations positively impact employees’ innovative passion. Based on the perspective of HRM system, the self-managed salary and job design in platform-based HRM practice can give employees a certain degree of autonomy [38], which can satisfy their autonomous needs; the team performance assessment can unite the team’s common goals and form the “shared identity” of the members [36], which can satisfy employees’ relational needs. The adaptive recruitment model can allow capable employees to enter the enterprise, provide them with development paths through career planning and with resources to achieve the purpose of empowerment so that they are more confident and capable of completing challenging work on their career paths, which satisfies their competency needs and stimulates their motivation to work. As platform organizations emphasize “openness, sharing, cooperation, and win–win” [1], the HRM model born in this organizational environment will promote the autonomous internalization of employees’ external motivation, and thus enhance employees’ harmonious passion for work. Based on this, the following hypothesis is proposed:

H1: 

Platform-based HRM practices have a significant positive impact on employees’ harmonious work passion.

In terms of platform-based HRM practices, firstly, autonomous job design allows employees to independently choose how to approach tasks and handle challenges within their roles, encouraging active participation. Self-managed compensation offers multiple pay options, enabling employees to select a scheme that best fits their circumstances, which enhances their sense of autonomy. Secondly, adaptive recruitment can select capable talents for the organization, and through empowering employee development, the personal qualities of employees can be comprehensively improved. At the same time, boosting development planning can clarify the future career development direction of employees so that they have more confidence and ability to complete the challenges of the work task and satisfy their competency needs. Finally, team-based performance management can motivate team members to commit to team tasks and rely on each other and endeavor to maintain team membership [39], which enhances trust and cohesion within the team [40], satisfying the employees’ need for belonging. Therefore, platform-based HRM practices satisfy all three needs of employees, which in turn enhances their sense of self-determination. Self-determination theory states that when the outside world can satisfy the three needs of employees, individuals will show positive behaviors. Reference [35] pointed out that employees with a sense of self-determination will maintain proactive innovative behaviors and enhance their passion for harmonious innovation. When the platform-based HRM practices meet employees’ basic psychological needs to a higher degree, the higher the employees’ sense of self-determination, the more time and energy they will proactively invest in their work and maintain a strong willingness or motivation to work proactively. Based on this, the following hypothesis is proposed:

H2: 

Self-determination can act as a mediator variable, explaining the predictive mechanism by which platform-based HRM practices influence employees’ harmonious work passion.

4.2. Variable and Data

Platform-based Human Resource Management Practices (PHRM) adopts the scale developed independently by this paper, which contains 6 dimensions and 22 items, which include self-managed salary management, team-based performance management, autonomous job design, empowering employee development, boosting development planning, and adaptive employee recruitment.

The sense of self-determination (SSD) scale, as used by Reference [35], consists of 9 items, and based on the data from this research, the Cronbach’s α value of this scale is 0.890. The results of the confirmatory factor analysis show that χ²/df = 1.219, RMSEA = 0.021, CFI = 0.967, TLI = 0.966, and SRMR = 0.022, and all fit indices met the criteria.

The Harmonious Work Passion (HWP) adopts the scale developed by Reference [37], which is strictly in accordance with the steps of “translation-back-translation”. Adjustments are made to align with the Chinese context and the platform-based organizational setting. This scale includes 7 items, with a sample item being “my work reflects that I identify with myself”. The Cronbach’s α of this scale is 0.909, and the results of confirmatory factor analysis show that χ²/df = 1.421, RMSEA = 0.032, CFI = 0.945, TLI = 0.943, SRMR = 0.037, SRMR = 0.037, TLI = 0.943, and SRMR = 0.945. SRMR = 0.037, and each fit indices met the criteria.

The data were collected using a combination of online and offline methods, and 21 platform-based enterprises and 21 non-platform-based enterprises in Tianjin, Guangzhou, Hangzhou, Qingdao, Chengdu, and Jinan were selected as research subjects. The choice of this method was based on the following considerations. Firstly, it is not always the case that those who have implemented a platform-based HRM practice model are platform-based enterprises. Secondly, if all of them chose to implement a platform-based HRM practice, the statistically sufficient variability could not be guaranteed. HR department managers and regular employees from each enterprise were invited to complete electronic questionnaires, with each enterprise involving 3 managers and 5–10 employees. The managers completed the platform-based HRM practices scale questionnaire, while regular employees completed the self-determination and harmonious work passion scale questionnaire. To ensure corresponding data between enterprises and employees, each company was coded with a unique letter identifier. The questionnaires were distributed in 42 sets (each company is regarded as a set of questionnaires), and 126 questionnaires for managers and 473 questionnaires for employees were recovered. After eliminating invalid responses, 37 sets remained, each containing three valid manager responses and five valid employee responses, for a total of 111 manager responses and 435 employee responses. In the manager sample, 38.7% were female and 61.3% male; 26.4% were aged 20–30, 53.5% aged 30–40, and 20.1% were over 40. In the employee sample, 30.8% were female and 69.2% male; 47.6% were aged 20–30, 34.1% aged 30–40, and 18.3% were over 40.

4.3. Hypothesis Testing

As the platform-based HRM practices scale developed for this study is an organizational-level variable, while employee self-determination and harmonious work passion are individual-level variables, cross-level model testing was conducted using SPSS 25.0 and Mplus 7.4.

First, the data aggregation test. Aggregating the data of three managers from each enterprise into an organizational layer required consistency checking. Firstly, SPSS 25.0 was used to test the consistency of the data of PHRM. The results show that the mean value of Rwg of PHRM is 0.921, which is higher than 0.7; the value of ICC (1) is 0.298, which is higher than 0.12; and the value of ICC (2) is 0.761, which is higher than 0.7, which indicates that the intra-group consistency of the sample data is high and that the data at the individual level can be aggregated to the organizational level by extracting the mean [41].

Secondly, variable descriptive statistics. The mean, standard deviation, and correlation coefficient of each variable were calculated using SPSS 25.0 software (

Table 7. Variable means, standard deviations and correlation coefficients.

Variant	Mean	Standard Deviation	C1	C2	C3	C4	HWP	SSD	PHRM
Layer 1 variables
Gender (C1)	0.308	/
Age (C2)	34.67	5.223	0.024
Educational level (C3)	/	/	0.052	0.009
Years of work experience (C4)	10.40	2.573	0.032	0.310 **	−0.131 *
HWP	3.140	0.667	0.014	−0.082	0.265 **	−0.197 *	(0.767)
SSD	2.998	0.775	0.061	0.008	0.045	−0.029	0.668 **	(0.878)
Layer 2 variables
PHRM	3.073	0.575	0.132	0.044	0.043	−0.061	0.559 **	0.709 **	(0.765)

Note: N1 = 435 (Stratum 1 variable), N2 = 37 (Stratum 2 variable), * p < 0.1, ** p < 0.05; C1 Employee gender, C2 Age, C3 Education, C4 Employee years of service.

). The correlation coefficient between the variables is under 0.6, which indicates that the sample data are reliable and the Common Method Bias of the data is not serious. Meanwhile, the correlation coefficient between the two variables is less than the square root of the AVE of each variable, which indicates that the discriminant validity of each variable meets the requirements and can be used as the data of this study.

Third, common method bias testing. Using Harman’s single-factor test analysis method, it was found that the first factor extracted from the unrotated exploratory factor analysis accounted for less than half of the total explanatory variables. This further indicates that the common method bias is not severe and that the survey data is representative.

Fourth, hypothesis testing. A cross-level model was constructed in Mplus 7.4 to examine the effect of PHRM on employees’ HWP. The results of the model are shown in

Table 8. Results of hypothesis testing analysis.

Relationship Path	Model 1
	Coefficient	Critical Ratio	95%CI
Within-Level Paths
C1→HWP	0.032	0.952	[−0.034, 0.099]
C2→HWP	−0.097	−2.853 **	[−0.163, −0.030]
C3→HWP	0.243	7.003 ***	[0.175, 0.312]
C4→HWP	−0.155	−4.461 **	[−0.223, −0.087]
SSD→HWP	0.543	8.363 **	[0.416, 0.671]
Cross-Level Paths
PHRM→SSD	0.779	30.651 ***	[0.748, 0.850]
PHRM→HWP	0.250	3.231 ***	[0.098, 0.402]

Note: ** p < 0.05, *** p < 0.01.

and Figure 3, from the results, it can be seen that PHRM practices have a cross-level positive effect on employees’ harmonious work passion, the direct effect (γ = 0.250, p = 0.000) and the total effect (γ = 0.673, p = 0.000) are both significant, supporting H1 and demonstrating the predictive validity of the scale, and the mediating effect is significant (γ = 0.423, p = 0.000), supporting H2 and indirectly validating the results derived from grounded theory.

Fifth, goodness-of-fit test. During the model calculation process, first, a model with only control variables and the explained variable as “HWP” was calculated. The goodness-of-fit R² of this model was 0.153. On this basis, after adding “PHRM”, the R² at Level 1 increased to 0.532 and the ΔR² was significant. Second, after adding the mediating variable “SSD”, the goodness-of-fit of the model was improved again, with the R² at Level 1 increasing to 0.701, and the ΔR² was significant. This indicates that the predictive effect model has strong explanatory power.

Based on the above cross-level model tests, it can be seen that the scale developed in this paper for platform-based human resource management practices can verify the predictive effects of such practices on self-determination and harmonious work passion. In practice, platform-based human resource management practices can be regarded as the input of an enterprise’s harmonious work passion—as a representative indicator of employees’ work motivation—which can be seen as the output of the enterprise. Employees’ self-determination represents the process mechanism between input and output. On the one hand, the research based on the above hypothesis indicates that this scale has good predictive validity; on the other hand, it also shows that the theoretical model derived from Figure 2 is highly consistent with real-world platform-based human resource management.

5. Discussion

5.1. Theoretical Contributions

(1) Examining the structural dimensions of platform-based HRM practices from a functional perspective. This study conceptualizes platform-based HRM practices as 6 dimensions: autonomous job design, empowering employee development, self-managed salary management, team-based performance management, boosting development planning, and adaptive employee recruitment. This dimensional structure shows that platform-based HRM practices are different from both control-based HRM and commitment-based HRM mentioned earlier, as well as inclusive HRM and developmental HRM proposed by scholars in recent years [11,42]. Platform-based HRM regards employees as partners [25], subverting the traditional perception of the relationship between organizations and employees. Conceptually, it transforms employees from “appendages” to partners and, in practice, provides them with equal entrepreneurial opportunities. By sharing benefits and assuming risks together, it turns employees into genuine “partners” [35], thus fully unleashing the creative value of individual employees in the digital era. Unlike Reference [10], our study divided platform-based HRM practices into platform empowerment, self-management, and ecosystem governance based on their effects on employees, this study focuses on HRM functions. Consequently, our findings provide more practical guidance for HRM practices.

(2) Development and validation of the platform-based HRM practices scale. This research developed a 22-item scale for platform-based HRM practices, covering 6 dimensions, and confirmed its reliability and validity. It also further expands the theoretical research on platform-based HRM practices, empirically demonstrating the positive effects of platform-based HRM practices on employees’ harmonious work passion, as well as the mediating role of the sense of self-determination. Existing studies have proposed the concept of platform-based HRM [9] and have made initial attempts to construct new HRM models for platform-based organizations [5]. Although specific measures for recruitment, training, utilization, and retention have been suggested, they have neither extracted a comprehensive HRM practice model nor developed a systematic measurement scale [25]. This study clarifies the dimensions and measurement indicators of platform-based HRM practices through grounded theory and draws on prior research, providing a quantifiable tool for future studies. It also expands the application scenarios of platform-based HRM practices and provides some references for future research on HRM practices in platform-based organizations.

(3) Clarify the mechanism of employee self-drive within platform-based HRM practices. The value of this study also lies in proposing the “platform empowerment–self-determination–work motivation” mechanism under platform-based HRM practices. This mechanism suggests that platform-based HRM practices not only meet employees’ psychological needs but also foster alignment between organizational and employee goals, jointly enhancing employees’ intrinsic motivation, which subsequently influences their behavior and motivation, providing the organization with sustained endogenous power. Different from the existing research [6,11,26], the platform-based HRM practice breaks through the previous role of “monitoring and management” and replaces it with the function of “service and incubation”, which treats employees as “business partners” and “empowers” them through the platform to achieve the purpose of employee self-management. The ultimate goal is to cultivate equal partners of the organization and bring lasting endogenous power to the organization. This conclusion is consistent with the findings of Reference [35], and it partially explains why the platform transformation of organizations enhances employee motivation and performance [1,8].

5.2. Practical Implications

(1) Enterprises need to re-examine the role positioning of organizational employee relations and human resource management and pay attention to the individual value of employees. In the industrial era, individuals depended on employment for livelihood, and capital relies on employment reproduction; in the digital economy era, the relationship between employees and organizations has evolved into an equal and cooperative relationship. As emphasized by the Chinese company Huawei, employees, “the closest to the gunfire”, can respond swiftly to market changes. Consequently, HRM should shift from “employee supervision” to “employee service”. Therefore, as a collection of resources, enterprises should integrate resources to empower employees, so that employees in the front, “expanding the territory” without worry. Moreover, establishing shared incentive mechanisms, employees can receive corresponding benefits when they meet user value and thereby realize their own value. This tightly binds employee interests with enterprise development, turning employees into true partners of the company. In addition, with the corresponding organizational culture, strengthening the cultural leadership, the equal partnership between the organization and the employees and the platform empowerment concept is embedded in the organizational culture, which can continue to influence the employees’ behaviors and attitudes, enhance the cohesion of the employees, promote the endogenous drive of the employees to achieve self-management, and give full play to the value of the individual creation.

(2) Providing a reference model for HRM innovation in platform-based and traditional organizations undergoing platform transformation. As platform-based organizations continue to emerge, corresponding HRM models remain underdeveloped. The six-dimensional platform-based HRM practices model constructed in this study offers practical reference value. For example, the Chinese company Haier has achieved mutual growth between employees and the organization through platform-based human resource management practices, in recruitment, involve hiring departments or teams in the entire selection process to avoid the phenomenon of “people are not in the right place” and attract talent through multiple channels, building a HR ecosystem. In job design, companies should grant their employees full autonomy to tackle tasks according to their preferences within defined goals and guidelines. In employee development, we make full use of digital technology to build a virtual scene so that employees can have a sense of immersion and achieve interaction with the user, implement the “combination of training and combat,” and put the training in the actual project, which reduces the cost of training and improves the comprehensive quality of employees. In salary management, reducing fixed pay while increasing the proportion of variable pay incentivizes potential entrepreneurs to achieve a reasonable proportion of super profits to be shared with the staff and vice versa, as well as to share the business risk. In performance management, personnel decision-making, distribution authority, and decision-making powers are delegated to teams. Teams are also required to determine their own performance levels, decentralizing risk, and aligning responsibilities, rights, and interests within the team. This delegation is supported by appropriate constraint mechanisms to prevent misuse of power. For example, a critical incident system can be implemented to record any deviant behaviors by employees in their exercise of authority [10]. In employee development planning, a “talent matching, not competition” development mechanism is implemented, which de-emphasizes employee evaluations. Instead, it focuses on nurturing talent through practical projects, providing employees with ample opportunities for growth and development.

5.3. Limitations and Perspectives

Firstly, the impact of platform-based HRM on employees is long-term. Due to the time factor, it is not possible to carry out long-term tracking and comparison. In the future, we can explore the long-term time effects of the impact of platform-based HRM on employees from the longitudinal aspect. Secondly, although the rootedness theory has many advantages, in the use of the process, it is still not free from subjective influence. For the problem of incomplete research, for this reason, the follow-up study will continue to carry out in-depth research on this issue. Finally, the impacts of platform-based HRM are complex and multifaceted, and in this study, only employees’ sense of self-determination and harmonious work passion were selected to test the predictive validity of platform-based HRM. In the future, the relationship between platform-based HRM and other univariate or multivariate variables can be explored to expand their application conditions and further test the scientific validity and applicability of the platform-based HRM scale.

6. Conclusions

The aim of the article was to explore the changes and innovations in HRM models or practices under the trend of organizational platformization. Firstly, following grounded theory procedures, it is argued that the platformized HRM practices under the “platform empowerment–self-determination–work motivation” mechanism are composed of self-managed salary management, team-based performance management, autonomous job design, empowering employee development, boosting development planning and adaptive employee recruitment. Secondly, the article also developed a measurement scale for platform-based HRM practices, which contains 22 formal items, Exemplary items include “we have the freedom to decide the pace and content of our work”, “in terms of compensation management, it feels like we’re working for ourselves”, “the company provides sufficient recruitment authority to the hiring departments”, and so on. The reliability and validity of this scale have been empirically confirmed. Finally, using harmonious work passion as a predictive indicator, cross-level model analysis reveals a significant positive impact of platform-based HRM practices on employees’ harmonious work passion, with self-determination serving as a mediator, supporting the predictive effectiveness of the platform-based HRM practices scale.

As shown in Figure 1, through the research of this paper, the structural dimension of platform-based human resource management practice was analyzed from the perspective of management function, and the scale of platform-based human resource management practice is developed and tested. On this basis, through the prediction effect test, it is found that the platform-based human resource management practice enhances the employees’ sense of self-determination and their enthusiasm for harmonious work, so as to clarify the mechanism of employees’ self-motivation under the platform-based human resource management practice model.