The construction sector has a set of very specific and unique characteristics that demarcate it from all other sectors. This is associated with a strong precari-ousness and labor turnover, plus the widespread practice of subcontracting. The pertinence of the study carried out is framed in the necessity of evaluating and valuing the management indicators, through accident rates that are statistically valid. In order to obtain the indices as a function of sectoral and market management variables, several statistical models were developed with the capacity to predict the behavior of these indices. Exogenous variables, such as the unemployment rate and the GDP growth rate of Portugal, were included for this purpose. These statistical models function as predictive models, provided that the coefficients of the independent variables are significant. Linear regressions were used because, due to the temporal shortage of data, the other types of regression would hardly prove to be robust. Also, were tested dozens of models, and the overwhelming majority did not show any statistical significance. Nevertheless, we obtained 3 (three) partially significant multiple linear regression models. Thus, as a final result of this work, it was verified that only (2) two models, demonstrated to have good predictability and reliability for future use. Being that these appear as relevant variables, the unemployment rate and the training costs per worker. These models can be used by companies in the industry, such as tools for the prevention of accidents at work.
CITATION STYLE
Oliveira, P., Baptista, J., & Pais, R. (2018). Proposal of a mathematical model of prediction of sinistrality values for valuation of organizational management indicators, applied to the construction industry. In Advances in Intelligent Systems and Computing (Vol. 604, pp. 375–383). Springer Verlag. https://doi.org/10.1007/978-3-319-60525-8_39
Mendeley helps you to discover research relevant for your work.