Bravo, Jorge Miguel VenturaLorente Salmerón, Antonio2020-03-132020-03-132020-02-27http://hdl.handle.net/10362/94203Presented as partial requirement for obtaining the Master’s degree in Statistics and Information Management, with a specialization in Information Analysis and ManagementHandling large amounts of granular data of non-financial corporations’ balance sheet and profit and loss statement to fulfil the assigned functions of the Central Balance Sheet Data Office (CBSDO) requires the design and implementation of various layers of quality control that combined with the review of expert analysts to ensure databases with a statistical quality at the expected height of the Statistics Department of an institution such as Banco de España. Recently, under the umbrella of the quality control systems of the CBSDO, a new outlier detection system has been designed and implemented, the “Outsider’s method”, through which those observations that deviate considerably from the behaviour of companies regarding sector clustering, according to NACE classification, and size, will be eliminated from database and will maintain those that still have a behaviour that is far from standard but consistent with the reality of mentioned Nonfinancial corporations. The purpose of this project is to explain the methodology of this new system, obtain results for different extractions periodically established over several years, analyse these results, and finally test the system's validity by comparing it with other detection methods traditionally used by other statistical entities.engNon-financial corporationsCBSDOOutlierQuality controlClusteringmaster thesis202456730