Uso de dados documentais para previsão e comparação dos custos da construção civil nas diferentes regiões do Brasil

Abstract

Being an area in constant growth and with great participation in the Brazilian GDP, the civil construction is a macro sector that involves several fields and industrial segments. In times when the economy of the country is uncertain, it becomes imperative to use statistical tools applied to forecasting costs before risking an investment. Therefore, the objective of this study is to identify the Brazilian region that offers the greatest cost benefit of the work through the forecast, considering three factors: cost of square meter, cost of labor persquare meter and cost of material per square meter. Initially, a theoretical study was carried out on the subject and a series of historical data on the proposed variables of the 5 main regions of Brazil was collected through IBGE. Next, with the help of Microsoft Office Excel, Action Stat Pro and Past software, a series of descriptive analysis of the series was made and the costs of 2017 were predicted. In order to verify the accuracy of historical data, it was decided to perform the MAP, MAD and MSD methods of exponential smoothing, verifying the quality of the samples and later, the quality of the predictions using the Theil’s U method. When comparing the results, this study provided current information on the monetary variations of the sector, obtaining as a response the Northeast region as the cheapest to be invested in the series studied, providing knowledge of the scope of construction with information on some of the main costs involved this field.