Science Research Management

Science Research Management ›› 2021, Vol. 42 ›› Issue (8): 150-159.

Previous Articles     Next Articles

A research on the influence factor and path of regional innovation based on the spatial structure equation model

Xue Yonggang   

  1. Medical Business School, Guangdong Pharmaceutical University, Guangzhou 510006, Guangdong, China
  • Received:2019-01-21 Revised:2019-12-29 Online:2021-08-20 Published:2021-08-13

Abstract:      Innovation is considered a key driver to economic growth and competitiveness. Traditional industrial economies are now transformed into knowledge economies where innovation is considered one of the mains drivers for sustained economic growth. Therefore, most of the advanced industrialized nations are keen to build their regional innovation systems. It is necessary to strengthen the scientific and technological innovation and to research regional science and technology innovation strategies. It is particularly important to apply scientific methods to study regional innovation now. This paper attempts to research regional innovation system of Guangdong province based on the exploratory spatial data analysis and the spatial structure equation model.
    We find that most of the domestic and foreign literature focuses on the following four aspects: (1) research on the influencing factors of regional innovation; (2) research on the spatial characteristics of regional innovation; (3) research on the efficiency of regional innovation; (4) research on the influence of regional innovation on enterprises. The existing literature finds that technological innovation plays an important role in determining the foreign direct investment-economic growth relationship. The level of collaboration with different partners can enhance firms′ innovation capabilities only if the focal firms′ managers have developed the capacity to scan and acquire external knowledge. There are positive effects of internal R&D and externally sourced innovation practices, as well as a positive influence of organizational innovation on the realization of technological innovations. The efficiency of innovation between regions has obvious spatial linkage and dependence. There is a significant positive spatial spillover effect. Unrelated variety hinders regional innovation efficiency and spatial spillover effects are not significant.
    Regional innovation is a complex system with the interaction of multiple factors, and there exist spatial spillover effects in the regional innovation system. Therefore, it is necessary to study the regional innovation according to the interaction of multiple factors & spatial spillover effects. Our study differs from the existing studies in two aspects: (1) according to the interactive effects of multiple factors including the innovation subjects, elements, policy and economic environment, and (2) expanding the structure equation model to the spatial structural equation model based on the exploratory spatial data analysis.
    Firstly, this paper analyzes the relation of the innovation elements, innovation subjects, innovation policy, the external economic environment and the internal economic environment. Then nine hypotheses are proposed as following: Hypothesis 1: The innovation subjects have a positive effect on the regional innovation; Hypothesis 2: The innovation elements have a positive effect on the regional innovation; Hypothesis 3: The innovation policy have a positive effect on the regional innovation; Hypothesis 4: The innovation policy have a positive effect on the innovation elements; Hypothesis 5: The innovation policy have a positive effect on the innovation subjects; Hypothesis 6: The internal economic environment has a positive effect on the regional innovation; Hypothesis 7: The internal economic environment has a positive effect on the innovation elements; Hypothesis 8: The internal economic environment has a positive effect on the innovation subjects; Hypothesis 9: The external economic environment has an inconclusive effect on the regional innovation. The Structure Equation Model of the regional innovation system is established based on the theoretical analysis and above 9 hypotheses.
    The exploratory spatial data analysis and structural equation model are adopted in the paper. Spatial analysis is statistically important because it can enhance inference accuracy, and at the same time reduces estimate bias by considering spatial proximity and dependence. The exploratory spatial data analysis should be considered as a descriptive step to explain the spatial patterns understudy and before estimating and testing more sophisticated regression models. The exploratory spatial data analysis can reveal complex spatial phenomena. Global autocorrelation is assessed by global Moran′s I statistic. A positive Moran′s I value indicates a general pattern of clustering in space of similar values. Structural equation modeling has roots in two different techniques developed in two different fields. Path analysis with its graphical representations of effects and effect decomposition comes from genetics research. Structural equation modeling is a confirmatory technique, which means that a model is formulated based on the theory, and it is judged whether this model should be rejected by fitting the model to data. If multivariate normality of the data holds, the variance-covariance matrix of the interest and the sample size is sufficient to fit models to the data. In order to test hypotheses, these hypotheses have to be translated in the a statistical model. The statistical model can be formulated in different ways. The structural equation model is defined by two main parts: the first part-the Structural Model- refers to the relationships among the latent variables, while the second part-the Measurement Model- represents the relationships between manifest and latent, endogenous and exogenous variables, respectively.
   All the data in this paper comes from the Guangdong Statistical Yearbook, Guangdong Intellectual Property Yearbook and China City Statistical Yearbook. The data time range is from 2005-2016. The space character and relativity were analyzed based on the exploratory spatial data analysis, the influence factor and mechanism were researched based on the spatial structure equation model with the example of 21 prefecture-level cities in Guangdong. We can gain the following conclusions: (1) The innovation policy, innovation elements and the external economic environment have a positive effect on the innovation, the innovation policy has more effect than the others; (2) The innovation subjects have a negative effect on the innovation; (3) The internal economic environment has a positive effect on the innovation and shows negative effect according to interaction; (4) The space relation has a positive effect on innovation. 
    The following suggestions are recommended: (1) It should restrict the scale of foreign investment and improve the quality of foreign capital; (2) The innovation evaluation and incentive system in colleges and universities should make better and implement innovation evaluation system of diversification and marketization; (3) The public platform of innovation system should be built by the government in Guangdong and provide the function of information exchange, multi-party cooperation, achievement trading and demand release; (4) The investment in educational science and technology must be enhanced and to create the scope of innovation and enhance the training of innovative talents.

Key words:  regional innovation system, spatial structure equation model, influence factor, influence mechanism