China
Africa
Explore chapters and articles related to this topic
Capability Analysis of Nonnormal Data
Published in Neil W. Polhemus, Process Capability Analysis, 2017
While the relationship between θ and Z can be used to calculate point estimates of equivalent capability indices, it cannot reliably be used to calculate confidence limits. This is because the properties of the estimated parameters of other distributions are not as well studied as for the normal distribution. One possible solution to this problem is to use bootstrapping. Bootstrapping is a method by which the sampling variability of a statistic can be determined by repeatedly sampling the available data.
Adopting cross-laminated timber in architectural design to reduce embodied carbon emission in China based on the diffusion of innovation theory
Published in Building Research & Information, 2023
Path coefficients indicate hypothesized relationships that link structures (Hair Jr et al., 2014). Having tested the reliability and validity of the measurement model, it is necessary to estimate the significance of the path coefficients to test the hypotheses within the framework of the structural model. PLS-SEM does not assume that data are normally distributed. Thus, PLS-SEM implements nonparametric bootstrapping (Davison & Hinkley, 1997). Bootstrapping involves repeated random sampling, replacing the original sample to create a bootstrap sample to obtain standard errors for hypothesis testing. The bootstrap samples allow significance testing of coefficients estimated in PLS-SEM (Henseler et al., 2009). Consistent with recommendations, the size of the bootstrap subsample in the present study was 5000. critical t-values for the two-tailed test were 1.65 (significance level = 10%), 1.96 (significance level = 5%) and 2.58 (significance level = 1%) (Hair Jr et al., 2021). A t-value below 1.96 is not in line with the recommendations (Rampasso et al., 2019). Figure 2 and Table 5 show the results of the hypothesis tests for the proposed research model. The model was tested using four endogenous variables (TR, RA, RI and IC). Eight out of 13 hypotheses were supported. Among these hypotheses, H1b, H1c, H3a, H3b, H3c, H4, H5 and H6 were supported by the data from the survey, while H1a, H2a, H2b, H2c and H7 were rejected.
She has got a gig: affordances of on-demand work apps for marginalised women enduring time scarcity
Published in Behaviour & Information Technology, 2023
Alka Agarwal, Ashish Kumar Jha, Jyoti Jagasia
We ran bootstrapping with 5000 resamples (Henseler, Ringle, and Sarstedt 2015). Our model had an acceptable Standardized Root Mean Squared Residual (SRMR) score of 0.064. We checked that all the constructs had Variance Inflation Factor (VIF) below 3 (Hair, Risher, and Sarstedt 2019). Since collinearity was not found in our study, we examined the R2 value which measures the model’s explanatory power (Benitez et al. 2020). In our model, the R2 values of all the five affordances are explained adequately while the R2 value of the dependent variable ‘LS’ is 0.379. There is medium predictive relevance of endogenous variables measured using Stone-Geisser’s Q2 value at 0.192 (Hair, Risher, and Sarstedt 2019). These assessment results demonstrate the validity of our structural model.
Efficacy in simulating the peak discharge response using soft computing techniques in the Jhelum river basin, India
Published in International Journal of River Basin Management, 2020
Dar Himayoun, Farooq Mohsin, Thendiyath Roshni
The consistency of the hydrological forecasts is mainly due to the following source of uncertainties (Bates and Townley 1988): data uncertainties, uncertainty of model arrangement and uncertainty of variables. In order to remove these uncertainties, a recent trend of bootstrapping is used. Bootstrapping is a simulation method (data-driven) which performs intensive re-sampling of data with replacement in order to reduce the uncertainties in the given set of data. Efron (1979) was the first to develop the bootstrap method that manipulates the given data to develop different kinds of models and generate aggregated potential predictors from them. Tiwari and Chatterjee (2011) used this bootstrapping method to forecast an hourly flood and found that the bootstrapping method is efficient enough to measure the uncertainty in flood predictions.