China
Africa
Explore chapters and articles related to this topic
Optimization Mechanisms When Confronted with the Variability of Sectoral Performance Factors
Published in Walter Amedzro St-Hilaire, Value-Based Management in an Open Economy, 2023
The regression uses both OLS and panel models (FEs or RE models). However, OLS, and panel techniques can suffer from time-varying specific effects and endogenous and causality problems. The existence of endogeneity problems in OLS variables is examined here using the Durbin-Wu-Hausman test (Durbin, 1954; Hausman, 1978; Wu, 1973). The current tests, when using the performance variables as measured by Tobin’s Q, ROA, and ROI, fail to accept the null hypothesis HO: the explanatory variables are thus exogenous. In other words, our results indicate that the hypothesis of no endogeneity is rejected. These results confirm the fact that OLS models and panel techniques are unreliable and biased. Therefore, this chapter concludes that GMM is the most appropriate approach. The results of the DWH tests point to endogeneity problems, which is another reason to opt for GMM regression.
Economic Development, Environmental Degradation and Sustainability
Published in Uday Chatterjee, Arindam Biswas, Jenia Mukherjee, Dinabandhu Mahata, Sustainable Urbanism in Developing Countries, 2022
Nilendu Chatterjee, Bappaditya Koley, Anindita Nath, Uday Chatterjee
Here we have followed Newey (1985) and Smith and Blundell (1986) and used the Durbin–Wu–Hausman test and the Hansen test for estimation of the simultaneous equation system. We know that the endogeneity issue for all the three equations of our study is suggested by the Durbin–Wu–Hausman test. This test also helps us to understand if instrumental variables techniques are needed or not because a rejection of the null hypothesis implies the significant as well as meaningful impact of endogenous variables on the parameters. Then, we have used the Hansen test for finding the proper instruments to determine the over-identifying restrictions in order to test the instruments’ validity. The validity of instrument implies that a null hypothesis of over-identifying restrictions cannot be rejected; it also implies that the instruments are appropriate.
Latent Variable Models
Published in Simon Washington, Matthew Karlaftis, Fred Mannering, Panagiotis Anastasopoulos, Statistical and Econometric Methods for Transportation Data Analysis, 2020
Simon Washington, Matthew Karlaftis, Fred Mannering, Panagiotis Anastasopoulos
This chapter presents tools for illuminating structure in data in the presence of measurement difficulties, endogeneity, and unobservable or latent variables. Structure in data refers to relationships between variables in the data, including direct or causal relationships, indirect or mediated relationships, associations, and the role of errors of measurement in the models. Measurement difficulties include challenges encountered when an unobservable or latent construct is indirectly measured through exogenous observable variables. Examples of latent variables in transportation include attitudes toward transportation policies or programs such as gas taxes, van pooling, high occupancy vehicle lanes, and mass transit. Interest might also be centered on the role of education, socio-economic status, or attitudes on driver decision making, which are also difficult to measure directly. As discussed in previous chapters, endogeneity refers to variables whose values are largely determined by factors or variables included in the statistical model or system.
Circular economy practices and corporate social responsibility performance: the role of sense-giving
Published in International Journal of Logistics Research and Applications, 2023
Tao Hong, Jinghua Ou, Fu Jia, Lujie Chen, Ying Yang
Examining the effect of long-term implementation of ECO (RL) practices on CSR performance may be difficult due to endogeneity concerns, including the omitted variable bias, sample selection bias, and two-way causality. For omitted variable bias, We use a balanced panel dataset and a fixed-effects model to capture any unobserved heterogeneity and unobserved time effects. We then control as many variables as possible, including firm-level and industry-level variables. In the sample screening process, we eliminate a large number of firms due to missing data on CSR scores, causing a sample selection bias concern. We thus use a standard Heckman (1979) two-step procedure which allows us to control for the unobservable factors that make sample inclusion more likely. The two-way causality is likely to occur because firms with higher CSR performance may obtain more financing to invest in CE (Cheng, Loannou, and Serafeim 2014). The instrumental variable (IV) method is used to further address this endogeneity concern. In addition, the panel Tobit model is adopted to reduce estimation inconsistency caused by the limited dependent variable with a value between 0 and 100 (Yi et al. 2020). Several alternative measurements of independent variables are also employed to perform further robustness tests.
A conceptualization of the spatial relationship associated with school-related crashes: a case study in Northwest Florida
Published in Transportation Planning and Technology, 2023
Mohammadreza Koloushani, Mahyar Ghorbanzadeh, Eren Erman Ozguven, Alireza Ermagun
Based on the research gap identified in previous studies, this study aspires to conduct a comprehensive spatial statistical analysis and shed more light on school-age children-involved crashes from different viewpoints. Unobserved heterogeneity, as an inevitable outcome of deficiency in crash data, has been widely explored in previous research and several advanced statistical methods have been proposed to deal with this issue (F. L. Mannering, Shankar, and Bhat 2016). Conversely, endogeneity refers to a situation in which an unidentified predictor variable is correlated with the error term in the regression model (F. Mannering et al. 2020). One of the most representative sources of endogeneity is omitted variables that could affect both response and indicator variables. Since most of the previous research specifically focused on school zones (Park, Abdel-Aty, and Lee 2019; Rahman et al. 2019), they did not examine the possible contribution of the spatial location of the schools compared to the most facilitated areas in the cities, which are Central Business Districts (CBDs). Moreover, an omitted variable may result in biased or incorrect findings. A thorough literature review reveals the gap in existing knowledge regarding the above-mentioned possible factor that may have a significant correlation with school-related crashes. This study intends to identify spatial factors that may influence this specific type of crash, either directly or indirectly, in terms of severity and frequency focusing on Leon County, Florida.
The impact of Operations and IT-related Industry 4.0 key technologies on organizational resilience
Published in Production Planning & Control, 2022
Giulio Marcucci, Sara Antomarioni, Filippo Emanuele Ciarapica, Maurizio Bevilacqua
According to several authors, the firm's dimension is relevant in influencing the impact level of Industry 4.0 projects. In general, smaller firms are more likely to experience difficulties in innovative activities (Müller, Buliga, and Voigt 2018; Dalenogare et al. 2018). Through an endogeneity bias test, the firm's dimension influence is tested, both in terms of number of employees and turnover. The procedure requires adding control variables useful to understand the study's validity: two extended models are built considering the number of employees and the firms' turnover as control variables. Indeed, endogeneity bias concerns the impossibility to interpret the effect of a predictor on a dependent variable because a variable that simultaneously affects both the predictor and the dependent variable is omitted in the model (Antonakis et al. 2014).