Small and medium-sized enterprise digital transformation method based on SEM and fsQCA

Abstract

The invention discloses a small and medium-sized enterprise digital transformation method based on SEM and fsQCA. The method comprises the following steps: obtaining samples and data of a target industry; a quantitative method is adopted through questionnaire survey, and variables are quantified by adopting a Like five-point scale; symmetrical and asymmetrical modeling is carried out on the measuring scale; carrying out comprehensive inspection and evaluation analysis on the proposed model; the enterprise digital transformation method has the beneficial effects that the organization phenomenon is understood in detail by integrating symmetric and asymmetric statistical tools, that is, SEM and fsQCA methods are integrated, so that small and medium-sized enterprises are helped to realize digital transformation, the possibility and interaction of organizations can be deeply informed, and the transformation efficiency is improved. The inherent limitation related to a symmetric method is solved, small and medium-sized enterprises, policy formulators and employees are guided to promote sustainable development practice by using digital technology, dynamic ability and a digital platform, and transition of the small and medium-sized enterprises from a simulation process to a digital process is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise digital management, and in particular to a digital transformation method for small and medium-sized enterprises based on SEM and fsQCA. Background Art

[0002] In the existing technology, with the development of informatization, small and medium-sized enterprises are facing unprecedented competitive pressure. The traditional business model can no longer meet the rapid changes in the modern market and the diversification of customer needs. Digital transformation has become a key path for small and medium-sized enterprises to enhance their competitiveness and achieve sustainable development. Digitalization can not only improve the operational efficiency of enterprises, but also optimize customer experience and enhance market response speed, thereby occupying a favorable position in the fierce market competition. Digital transformation is inseparable from advanced technical support. Emerging technologies such as cloud computing, big data, the Internet of Things, and artificial intelligence provide small and medium-sized enterprises with a variety of tools and platforms to help them build intelligent production, management, and service systems. For small and medium-sized enterprises, digital transformation is not only an inevitable choice to enhance corporate competitiveness, but also an important way to adapt to market changes and achieve sustainable development. Small and medium-sized enterprises also face many challenges in the process of digital transformation, especially how small and medium-sized enterprises can use digital platforms for digital transformation.

[0003] The Chinese invention patent with application number: 202411627438.2 discloses a method and system for recommending solutions for digital transformation of small and medium-sized enterprises, the method comprising: obtaining business data between multiple business nodes and multiple business nodes in a historical time period; determining the time variation characteristics of the business data between multiple business nodes according to the business data between multiple business nodes in a historical time period, and determining the digital transformation solutions between multiple business nodes according to the business data between multiple business nodes in a historical time period and the time variation characteristics of the business data between multiple business nodes, determining multiple first business nodes whose fluctuation values ​​of the data volume of multiple data types are less than a preset fluctuation value, and determining multiple second business nodes whose fluctuation values ​​of the data volume of multiple data types are greater than or equal to the preset fluctuation value. However, this method and system for recommending solutions for digital transformation of small and medium-sized enterprises can only guide the transformation direction of enterprises, but it does not disclose how small and medium-sized enterprises can promote their transition from analog processes to digital processes by adopting new technologies, software or digital platforms to carry out business, and it does not disclose how to build a model to promote digital innovation in the sustainable development paradigm and how small and medium-sized enterprises can carry out digital transformation based on: partial least squares structural equation modeling (SEM) and fuzzy set qualitative comparative analysis (fsQCA). Summary of the invention

[0004] The purpose of this invention is to provide a digital transformation method for small and medium-sized enterprises based on SEM and fsQCA.

[0005] To achieve the above object, the technical solution proposed by the present invention is as follows:

[0006] A digital transformation method for small and medium-sized enterprises based on SEM and fsQCA, comprising the following steps:

[0007] Obtain samples and data of the target industry;

[0008] Adopt a quantitative method through questionnaire surveys, and quantify variables using a Likert five-point scale;

[0009] Perform symmetric and asymmetric modeling on the measurement scale;

[0010] Conduct a comprehensive inspection and evaluation analysis of the proposed model;

[0011] Summarize the digital transformation method of enterprises.

[0012] Obtain samples and data of the target industry, including:

[0013] The research samples are small and medium-sized enterprises in the target industry;

[0014] Adopt a stratified random method to select samples;

[0015] Select the research objects of the research group, and the research objects are middle and senior management personnel of some small and medium-sized enterprises;

[0016] Through cooperation with the target industry association, distribute questionnaires, collect the questionnaires and eliminate invalid questionnaires.

[0017] Adopt a quantitative method through questionnaire surveys, and quantify variables using a Likert five-point scale, including:

[0018] Adapt the items of existing research to ensure relevance and appropriateness;

[0019] Quantify the variables among them using a Likert five-point scale.

[0020] Perform symmetric and asymmetric modeling on the measurement scale, including:

[0021] For symmetric modeling, select partial least squares structural equation modeling, i.e., PLS-SEM;

[0022] For asymmetric modeling, select fuzzy set qualitative comparative analysis, i.e., fsQCA.

[0023] Conduct a comprehensive inspection and evaluation analysis of the proposed model, including:

[0024] Conduct a comprehensive inspection of the measurement model and evaluate the measurement mode;

[0025] Evaluate the structural model and the predictive ability of the model;

[0026] Qualitative Comparative Analysis of Fuzzy Sets

[0027] When comprehensively examining the measurement model, its internal consistency is measured by Cronbach's coefficient and composite reliability statistic, and the convergent validity is evaluated by the average variance extraction method;

[0028] When evaluating the measurement model, the HTMT correlation ratio method and the Fornell-Larcker criterion are used to confirm the discriminant validity to evaluate the differences between each potential construct and other constructs.

[0029] Structural model evaluation and the predictive ability of the model, including:

[0030] The consistency between the sample data and the proposed co-adjustment model is analyzed by SEM, and the overall evaluation results are shown by the fit indices;

[0031] Through the coefficient of determination R 2 and the predictive relevance Q (2) The predictive ability of the model is evaluated.

[0032] Qualitative Comparative Analysis of Fuzzy Sets, including:

[0033] Data calibration;

[0034] Necessary condition analysis;

[0035] Configuration analysis.

[0036] Data calibration involves transforming individual variables into coherent conceptual categories and then assigning cases to these collective categories;

[0037] Adopt a combinatorial perspective to grasp dynamic changes for necessary condition analysis;

[0038] Specific criteria for case frequency and raw consistency threshold are established, set to 1 and 0.80 respectively, and the proportion reduction in inconsistency, i.e., PRI, consistency index is used as a filtering mechanism for the truth table, with the threshold set to 0.70 for configuration analysis.

[0039] Summarize the enterprise digital transformation methods, including:

[0040] Discuss the analysis results and summarize the digital transformation methods.

[0041] The beneficial effects of the present invention are:

[0042] By integrating symmetric and asymmetric statistical tools, namely integrating the SEM and fsQCA methods, to gain a more nuanced understanding of organizational phenomena to assist small and medium-sized enterprises in achieving digital transformation, it is possible to gain in-depth insights into organizational possibilities and interactions, address the inherent limitations associated with symmetric methods, guide small and medium-sized enterprises, policymakers, and practitioners in leveraging digital technologies, dynamic capabilities, and digital platforms to promote sustainable development practices, facilitate the transition of small and medium-sized enterprises from analog processes to digital processes, and guide small and medium-sized enterprises in constructing a model that promotes digital innovation within the sustainable development paradigm. Description of the Drawings

[0043] Figure 1 is a schematic diagram of the overall structure of the present invention;

[0044] Figure 2 is a measurement table of the variables of the present invention;

[0045] Figure 3 is an evaluation table of the measurement model of the present invention;

[0046] Figure 4 is a discriminant validity table of the present invention;

[0047] Figure 5 is a direct path analysis table of the present invention;

[0048] Figure 6 is a structural model diagram of the present invention;

[0049] Figure 7 is an indirect path (mediation) analysis table of the present invention;

[0050] Figure 8 is a moderation analysis table of the present invention;

[0051] Figure 9 is a moderation effect diagram of the present invention;

[0052] Figure 10 is a necessary condition analysis diagram of the present invention;

[0053] Figure 11 is a configurational analysis table of the SDI of the present invention. Detailed Description of the Invention

[0054] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0055] A method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA, comprising the following steps:

[0056] Obtain samples and data of the target industry;

[0057] Adopt a quantitative method through a questionnaire survey, and the variables are quantified using a Likert five-point scale.

[0058] using symmetric and asymmetric modeling for measurement scales;

[0059] Conduct a comprehensive inspection and evaluation analysis of the proposed model;

[0060] Summarize the enterprise digital transformation methods.

[0061] Among them, the schematic diagram of the transformation method of the present invention is as follows Figure 1 shown.

[0062] Obtain samples and data on target industries, including:

[0063] The research sample is small and medium-sized enterprises in the target industry;

[0064] The sample was selected using a stratified random method;

[0065] The research group selected the subjects of the study, which were middle and senior managers of some small and medium-sized enterprises;

[0066] By cooperating with the target industry associations, we distributed questionnaires, collected them back and eliminated invalid questionnaires.

[0067] Among them, the research samples of this embodiment include small and medium-sized enterprises in industries such as mining, papermaking and textiles. These small and medium-sized enterprises have participated in digital transformation to varying degrees and have different dynamic capabilities. Each selected manager plays an important decision-making role in the formulation, selection and implementation of the company's sustainable strategy, reflecting the overall sustainable cognition level of the company. The sample selection is based on the classification of polluting industries such as printing, petroleum, electricity, and heat in the "List of Environmental Information Disclosure of Listed Companies". Data collection includes on-site surveys and online questionnaires. Through cooperation with industry associations, a total of 520 questionnaires were issued and 434 were collected. After eliminating 51 invalid questionnaires, 386 valid questionnaires were collected, with an effective recovery rate of 74.23%.

[0068] This embodiment explores the digital transformation methods of SMEs through four dimensions: digital technology, dynamic capabilities, digital platform DP and digital divide DD. Digital technology includes artificial intelligence, material network, mobile technology MT and blockchain BT. Dynamic capabilities include perception capability SC, grasp capability ZC, innovation capability IC and reconstruction capability RC. Digital divide DD includes acquisition, resources and power. Through the above dimensions and related variables, SEM and fsQCA methods are used to explore the digital transformation methods of SMEs to achieve sustainable digital innovation SDI.

[0069] A quantitative method was used through a questionnaire survey, and the variables were quantified using a five-point Likert scale, including:

[0070] Items from existing studies were adapted to ensure relevance and appropriateness;

[0071] The variables among them are quantified using a Likert five - point scale.

[0072] Among them, a detailed introduction to the variables and their related measurement items is as Figure 2 shown.

[0073] Symmetric and asymmetric modeling are adopted for the measurement scale, including:

[0074] For symmetric modeling, partial least squares structural equation modeling, i.e., PLS - SEM, is selected;

[0075] For asymmetric modeling, fuzzy - set qualitative comparative analysis, i.e., fsQCA, is selected.

[0076] Among them, the symmetric analysis method is based on the idea that the level of the predictor is usually associated with the corresponding low or high value of the outcome. The asymmetric model shows that a high value of the predictor is often a sufficient condition for the high importance of the criterion variable to occur, but not necessarily a mandatory condition. This model shows that even when the value of the predictor variable is low, the level of the criterion variable may increase.

[0077] PLS - SEM is used to perform symmetric analysis and can effectively handle the complexity of moderation models and mediation models. During the execution of PLS - SEM, a systematic method is adopted. First, a thorough examination of the measurement scale is required to ensure the reliability and validity of the constructs under study. Subsequently, SEM analysis is used to evaluate the hypothesized relationships in the model. As a symmetric analysis method, PLS - SEM is good at evaluating the cumulative net effect of predictor factors on the overall result of the model.

[0078] Using fsQCA as the analysis method can clarify intricate causal relationships, especially regarding the combination of predictor factors that lead to a specific outcome, and can clarify how different factors are consistent with the combination of necessary and sufficient conditions that constitute a specific outcome.

[0079] A comprehensive inspection and evaluation analysis is carried out on the proposed model, including:

[0080] Conduct a comprehensive inspection of the measurement model and evaluate the measurement pattern;

[0081] Evaluate the structural model and the predictive ability of the model;

[0082] Fuzzy - set qualitative comparative analysis.

[0083] Among them, to detect common method bias, i.e., CMB, a common method factor is added to the model. The analysis results show that the method variance R2 2It is not obvious that the method factor loading R2 is not obvious, indicating the absence of CMB. The collinearity test confirmed that the VIF values, i.e., variance inflation factors, of all constructs are below 3.3. For the evaluation of the measurement model, please refer to the appendix of the specification Figure 3 , thus confirming that there is no threat to the validity of CMB.

[0084] When comprehensively examining the measurement model, its internal consistency is measured by Cronbach's alpha coefficient and composite reliability statistic, and the convergent validity is evaluated by the average variance extracted method;

[0085] When evaluating the measurement model, the HTMT correlation ratio method and the Fornell-Larcker criterion are used to confirm the discriminant validity to evaluate the differences between each latent construct and other constructs.

[0086] Among them, the comprehensive examination focuses on two aspects: reliability and validity. The evaluation of index reliability includes a careful examination of index loadings, and the internal consistency reliability is measured by Cronbach's alpha coefficient and composite reliability, i.e., the CR statistic. When these reliability indicators exceed the critical value of 0.70, they are considered satisfactory. The convergent validity is evaluated by the average variance extracted method, i.e., AVE. An AVE value exceeding 0.70 is considered satisfactory, and 0.50 is considered an acceptable indicator. According to the empirical rule for detecting potential multicollinearity problems, the VIF value needs to be checked. A threshold of 3.3 or higher indicates the presence of multicollinearity.

[0087] Discriminant validity is used to evaluate the differences between each latent construct and other constructs, and it is required that the AVE value of each construct must exceed the square of the correlation between this construct and any other latent construct.

[0088] For the HTMT correlation ratio method and the Fornell-Larcker criterion, please refer to the appendix of the specification Figure 4 , the HTMT method includes checking the HTMT value. If the HTMT value is close to or exceeds 1, it indicates a lack of discriminant validity. When the square root of the AVE of each construct exceeds its correlation with any other construct in the model, the Fornell-Marcker criterion confirms the discriminant validity. When the diagonal elements exceed the other non-diagonal features in the row and column, this condition is met. The research results in the appendix of the specification Figure 4 show that the diagonal values are in bold, representing the square root of the AVE of each latent construct, affirming their superiority over the correlations of other constructs in the model.

[0089] The evaluation of the structural model and the predictive ability of the model include:

[0090] The consistency between the sample data and the proposed common adjustment model is analyzed through SEM, and the overall evaluation results are indicated by the fit indices;

[0091] By the coefficient of determination R 2 and the prediction correlation Q (2) Evaluate the prediction ability of the model.

[0092] Among them, the overall evaluation of the goodness-of-fit index shows that the co-alignment model has a high goodness-of-fit with the data, namely χ2: 402.71; CFI: 0.914; GFI: 0.952; RMSEA: 0.039, R 2 value indicates the degree of explanation of the SDI variation by the model, and this model accounts for 82.1% of the SDI variance, R 2 value exceeding 0.10 is acceptable, and this embodiment has a quite high R 2 value. In addition, after evaluating the prediction correlation using the Q 2 value, it is found that the SDI value is 0.318, and these positive values confirm that the model has a strong prediction correlation.

[0093] Direct effect analysis:

[0094] The results of the main effect analysis show that AI, β: 0.150, p≤0.05, BT, β: 0.263, p≤0.01, MT, β: 0.206, p≤0.01, ZC, β: 0.149, p≤0.05, IC, β:.153, p≤0.05 and RC, β: 0.149, p≤0.05 have a positive impact on DP.

[0095] Similarly, AI, β: 0.202, p≤0.05, BT, β: 0.206, p≤0.01, MT, β: 0.204, p≤0.01, ZC, β: 0.101, p≤0.01, IC, β: 0.140, p≤0.01, RC, β: 0.229, p≤0.01, SC, β: 0.160, p≤0.05 and DP, β: 0.083, p≤0.01 have a positive impact on SDI.

[0096] On the contrary, the Internet of Things, β: 0.077, p>0.05 and SC, β: 0.007, p>0.05 have no significant impact on DP, and the impact of the Internet of Things on SDI is also not significant, β: 0.054, p>0.05. For detailed conclusions, see the appendix of the specification Figure 5 and the appendix of the specification Figure 6 .

[0097] Mediation analysis:

[0098] For mediation analysis, see the appendix of the specification Figure 7, including significant indirect effects, DP on AI, β: 0.124, p ≤ 0.01, BT, β: 0.218, p ≤ 0.01, MT, β: 0.252, p ≤ 0.01, ZC, β: 0.112, p ≤ 0.01, IC, β:.043, p 0.01 and RC, β: 0.124, p ≤ 0.05. The mediating level was evaluated using the variance proportion VAF score. A VAF index value exceeding 80% indicates complete mediation, between 20% and 80% indicates partial mediation, and below 20% indicates no mediation. The research results show that DP exhibits partial mediation between the predictor and the outcome.

[0099] Moderation analysis:

[0100] For moderation analysis, see the appendix of the specification Figure 8 , and the results determined the moderating effect of DD. DD has a significant statistical effect on SDI, β: 0.104, p ≤ 0.01. DD on AI and SDI, β: 0.444, p ≤ 0.01, BT and SDI β: 0.460, p ≤ 0.01, MT and SDI, β: 0.328, p ≤ 0.01, SC and SDI, β: 0.425, p ≤.01, IC and SDI, β: 0.275, p ≤ 0.01, and RC and SDI, β: 0.404, p ≤ 0.01. The appendix of the specification Figure 9 shows these significant moderating effects. DD does not have a significant moderating effect between IoT and SDI, β: 0.102, p > 0.05 and ZC and SDI, β: 0.108, p > 0.05.

[0101] Fuzzy-set qualitative comparative analysis, including:

[0102] Data calibration;

[0103] Necessary condition analysis;

[0104] Configuration analysis.

[0105] Among them, the impact of digital technologies and capabilities on SDI was studied through regression analysis to confirm the multi-faceted nature of digitalization in management practice, emphasizing the complexity of digitalization and requiring an integrated approach to understand causal dynamics. By combining fsQCA and SEM, the interactions between variables were comprehensively examined. The fsQCA method adopts an overall approach, treating each case as a unique combination of antecedent conditions. By analyzing multiple equivalent paths affecting SDI, it promotes the exploration of complex relationships. Through sufficiency analysis, these paths can be examined to help determine the interactions between variables.

[0106] Data calibration involves transforming individual variables into coherent conceptual categories and then assigning cases to these collective categories;

[0107] Adopt a combined perspective to grasp dynamic changes for necessary condition analysis;

[0108] Specific criteria for case frequency and raw consistency threshold were established, set at 1 and 0.80 respectively, and the proportion reduction in inconsistency, i.e., PRI, consistency index was used as a filtering mechanism for the truth table, with the threshold set at 0.70 for configurational analysis.

[0109] Among them, the calibration procedures for the result variables and condition variables in data calibration mainly involve descriptive statistics and well-founded theoretical and practical insights. The calibration strategy integrates three anchor points: full entry, crossover point, and full exit, which depend on key statistical parameters in the sample dataset, including the maximum value, average value, and minimum value, to ensure that the variables are effectively consistent with the conceptual framework of the research basis. For necessary condition analysis, see the appendix of the specification Figure 10 The consistency coefficients of all conditions are lower than the critical value of 0.90. Therefore, no single condition can be regarded as a necessary condition for SDI.

[0110] Configurational analysis:

[0111] Specific criteria for case frequency and raw consistency threshold were established, set at 1 and 0.80 respectively. The proportion reduction in inconsistency PRI, consistency index was used as a filtering mechanism for the truth table, with the threshold set at 0.70. For the configurational analysis of SDI, see the appendix of the specification Figure 11 Analysis yielded five different configurational paths related to high SDI levels. The calculated overall solution consistency was 0.85, exceeding the threshold of 0.80, and the coverage rate reached 0.81.

[0112] Summarize the methods of enterprise digital transformation, including:

[0113] Discuss the analysis results and summarize the digital transformation process.

[0114] Among them, Configuration 1 shows that in a sustainable environment, the presence or absence of the Internet of Things and integrated circuits does not have a significant statistical impact on the generation of SDI. The key determining factors include the presence of the core conditions of AI, SC, and DD, which are associated with the presence of the peripheral conditions of BT, MT, ZC, and RC.

[0115] Configuration 2 shows that the presence of ZC has no significant statistical impact on SDI. When the core conditions, especially AI, SC, and DD, are met, and at the same time the peripheral conditions IoT, BT, and RC are present, while IC and MT are still intermittent, a high SDI will be obtained.

[0116] Configuration 3 shows that the presence or absence of AI and IC has no obvious effect on SDI. On the contrary, when the core condition DD is satisfied simultaneously with the peripheral conditions ZC and RC, SDI will be manifested. At the same time, the peripheral conditions BT, MT, and SC, as well as the core condition IoT, are still absent.

[0117] Configuration 4 shows that the obvious presence or absence of BT and ZC has no significant effect on SDI. When the core conditions SC, IC, and DD, as well as the peripheral conditions AI, MT, and RC, are satisfied while IoT is absent, SDI increases.

[0118] Configuration 5 shows that the presence or absence of the Internet of Things, SC, and IC has no significant statistical effect on SDI. On the contrary, when the core condition DD is satisfied simultaneously with the peripheral conditions MT and ZC, while the peripheral conditions AI, BT, and RC are absent, high SDI can be achieved.

[0119] Therefore, the original coverage rate of Configuration 1 is the highest, reaching 0.57, which is the best solution to improve SDI, and it can be proved that different combinations of digital technology, dynamic capabilities, and the digital divide are related to the increase in SDI.

[0120] The results of the main effect analysis show that for the digital transformation of small and medium-sized enterprises, in terms of digital technology, artificial intelligence, blockchain, and mobile technology have a positive impact on the DP of polluting small and medium-sized enterprises. By using artificial intelligence technology, enterprises can enhance their digital presence and functions, thus achieving positive results in the development of digital platforms. The use of blockchain can improve the transparency, traceability, and security of various processes, which is particularly important for industries dealing with environmental issues. The integration of blockchain technology helps to develop more reliable and secure digital platforms for polluting small and medium-sized enterprises in China. The wide application of mobile technology improves accessibility and connectivity, enabling enterprises to reach a wider audience. Mobile devices facilitate communication, data sharing, and customer participation.

[0121] In terms of dynamic capabilities, the capabilities of seizing opportunities, integrating capabilities, and reconfiguring capabilities have a positive impact on the DP of polluting small and medium-sized enterprises. For example, adopting new technologies or responding to market changes greatly promotes the development of digital platforms for small and medium-sized enterprises. Seamlessly integrating various technologies and reconfiguring existing processes to adapt to the digital environment is crucial for small and medium-sized enterprises to successfully develop digital platforms. That is, digital transformation requires an overall approach that includes both technology integration and organizational flexibility.

[0122] It can be concluded that the digital transformation method for small and medium-sized enterprises is as follows: adopt digital technologies to build a digital platform, improve the efficiency and transparency of the enterprise to enhance the dynamic capabilities of the enterprise itself and the connectivity within the enterprise. Using the digital platform as a channel, according to the digital gap between the enterprise itself and leading enterprises, collect relevant data, optimize resource utilization, and implement innovation strategies, discover and implement innovative solutions to promote sustainable practices, that is, address the gap in access to digital technologies, allocate sufficient resources, and utilize external forces, and take targeted intervention measures to bridge the gap to narrow the digital divide, thus enabling the digital transformation of small and medium-sized enterprises. The digital platform can promote interdisciplinary cooperation, ensure the maximization of the benefits of technology, and transform the potential of advanced digital technologies into a viable strategy to drive sustainable innovation. Dynamic capabilities contribute to the initiation of innovative practices and form a cohesive digital ecosystem. In this transformation method, the digital divide is used as a moderating variable to help small and medium-sized enterprises in need of transformation to carry out digital transformation based on the experience of leading enterprises in the same industry.

[0123] Working principle:

[0124] This technical solution adopts a novel methodology that combines the structural equation model SEM and the fuzzy set qualitative analysis fsQCA to gain a more nuanced understanding of organizational phenomena, address the inherent limitations associated with symmetric methods, and thus improve it to determine the specific combinations of various factors that jointly contribute to the desired outcome, that is, utilize the transformative potential of artificial intelligence, blockchain, and mobile technologies to drive innovation in small and medium-sized enterprises that aligns with sustainable goals, strategically enhance the perception ability of small and medium-sized enterprises to monitor environmental factors, seize opportunities for sustainable development initiatives, integrate various technologies, and reconfigure processes, so as to be consistent with sustainable development goals, strategically utilize the digital platform to integrate and streamline the functions of artificial intelligence, BT, and MT with dynamic capabilities such as grasping, integrating, and reconfiguring capabilities, narrow the digital divide, promote inclusiveness, and provide small and medium-sized enterprises with the necessary resources to utilize digital technologies to promote sustainable practices and improve the effectiveness of technology adoption and organizational capabilities.

[0125] Meanwhile, in the context of the digital economy, digital technologies and dynamic capabilities have catalyzed the SDI process. While enhancing digital platforms and innovation capabilities, digital technologies and dynamic capabilities also take into account the digital divide issue. Enterprises can effectively manage rapid SDI by proficiently adopting digital technologies and platforms, accumulating dynamic digital capabilities, and addressing the digital divide problem. Digital technologies include a series of electronic tools, systems, and solutions that utilize digitalized information and processes to strengthen different aspects of enterprise operations and strategies. Dynamic capabilities are important drivers of SDI. Dynamic capabilities include the enterprise's adaptability and innovation capabilities, which help effectively respond to dynamic market conditions, thereby obtaining sustainable competitive advantages. It aims to enhance the enterprise's adaptability, innovation capabilities, and the reconfiguration capabilities of resources and processes, ensuring the enterprise achieves lasting competitiveness and success in the ever-changing business environment. The interaction of dynamic capabilities, the digital divide, and digital platforms has had a significant impact on organizational efficiency, costs, and productivity, ultimately supporting the realization of SDI.

[0126] The beneficial effects of the present invention are to achieve a more nuanced understanding of organizational phenomena by integrating symmetric and asymmetric statistical tools, namely integrating the SEM and fsQCA methods, to assist small and medium-sized enterprises in achieving digital transformation. It can provide in-depth insights into organizational possibilities and interactions, solve the inherent limitations associated with symmetric methods, guide small and medium-sized enterprises, policymakers, and practitioners to utilize digital technologies, dynamic capabilities, and digital platforms to promote sustainable development practices, facilitate the transition of small and medium-sized enterprises from analog processes to digital processes, and guide small and medium-sized enterprises to build a model that promotes digital innovation in the sustainable development paradigm.

[0127] The above has elaborated in detail on an embodiment of the present invention, but the content described is only the preferred embodiment of the present invention and cannot be considered as defining the scope of implementation of the present invention. All equivalent changes and improvements made in accordance with the scope of the application of the present invention shall still fall within the patent coverage scope of the present invention.

Claims

1. A digital transformation method for small and medium-sized enterprises based on SEM and fsQCA, characterized in that, It includes the following steps: Obtain samples and data of the target industry; Adopt a quantitative method through questionnaire surveys, and quantify variables using a Likert five-point scale; Conduct symmetric and asymmetric modeling on the measurement scale; Conduct a comprehensive inspection and evaluation analysis on the proposed model; Summarize the enterprise digital transformation method.

2. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 1, characterized in that Obtain samples and data of the target industry, including: The research samples are small and medium-sized enterprises in the target industry; Adopt a stratified random method to select samples; Select the research objects of the research group, which are the middle and senior management personnel of some small and medium-sized enterprises; Through cooperation with the target industry association, distribute questionnaires, collect the questionnaires and eliminate invalid questionnaires.

3. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 2, characterized in that, Adopt a quantitative method through questionnaire surveys, and quantify variables using a Likert five-point scale, including: Adapt the items of existing research to ensure relevance and appropriateness; Quantify the variables using a Likert five-point scale.

4. The digital transformation method for small and medium-sized enterprises based on SEM and fsQCA according to claim 3, characterized in that, Conduct symmetric and asymmetric modeling on the measurement scale, including: For symmetric modeling, select partial least squares structural equation modeling, i.e., PLS-SEM; For asymmetric modeling, select fuzzy set qualitative comparative analysis, i.e., fsQCA.

5. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 4, characterized in that, Conduct a comprehensive inspection and evaluation analysis on the proposed model, including: Conduct a comprehensive inspection on the measurement model and evaluate the measurement mode; Evaluate the structural model and the predictive ability of the model; Fuzzy set qualitative comparative analysis.

6. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 5, characterized in that, When conducting a comprehensive inspection on the measurement model, measure its internal consistency through Cronbach's coefficient and composite reliability statistics, and evaluate the convergent validity using the average variance extraction method; When evaluating the measurement mode, use the HTMT correlation ratio method and the Fornell-Larcker criterion to confirm the discriminant validity to evaluate the differences between each latent construct and other constructs.

7. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 6, characterized in that, Evaluate the structural model and the predictive ability of the model, including: Analyze the consistency between the sample data and the proposed common adjustment model through SEM, and show the overall evaluation results through the fit index; Evaluate the predictive ability of the model by the coefficient of determination R 2 and the prediction correlation Q (2) ​ 8. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 7, wherein Fuzzy set qualitative comparative analysis, including: Data calibration; Necessary condition analysis; Configuration analysis.

9. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 8, characterized in that, Data calibration involves transforming individual variables into coherent conceptual categories, and then assigning cases to these collective categories; Adopt a combinatorial perspective to grasp the dynamic changes for necessary condition analysis; Establish specific criteria for case frequency and raw consistency threshold, set them to 1 and 0.80 respectively, and adopt the proportion reduction in inconsistency, i.e., PRI, consistency index as the filtering mechanism for the truth table, set the threshold to 0.70, and conduct configuration analysis.

10. The method for digital transformation of small and medium-sized enterprises based on SEM and fsQCA according to claim 9, characterized in that, Summarize the enterprise digital transformation method, including: Discuss the analysis results and summarize the digital transformation method.

Images