Outlier Detection with Kernel Density Functions in Monitoring the ISTAT LFS Data Production Processes
Abstract
To calculate the thresholds that determine the normal operating region for a data production process, it is proposed a method based on a time-varying tolerance bounds. The proposal is applied on the Italian Labour Force Survey to report how this new decision rules impact on survey quality
DOI Code:
10.1285/i20705948v6n1p118
Keywords:
monitoring data production, AAPOR, time series analysis
References
Ornella GIAMBALVO, Statistical methods for data analysis,
Luciana QUATTROCIOCCHI, Survey monitoring [email protected]
Full Text: PDF