DATA VALIDATION Foreign Trade Statistics ESTP course Data Validation in the ESS
Validation process of Foreign trade statistics data In Latvia CSB Foreign trade statistics data collection and processing section is responsible for Foreign trade statistics data collecting (in case of Intrastat), data entry (in case of paper declaration), data processing and data validation. Intrastat and Extrastat data sets are processed separately, they are merged only at the last stage of validation – during analysis of aggregated data. Intrastat data validation Extarstat data validation
INTRASTAT data processing schemedata processing scheme Collecting of reports via CSB e - survey Collecting of reports Data Import in SQL database FTS data analysis and comparing Adjustments calculation FTS data dissemination Registration of reports in ISDMS Registration module Data analysis and processing for Adjustments calculation Calculating of statistical value for A type reports Paper form Intrastat reports data inputting in ISDMS Data entry module Electronically received Intrastat reports data import in ISDMS Data entry module Validation of correctness of codes and credibility checking in ISDMS Data validation module 14-15/05/2014
Intrastat data validation process Phases of validation process: Primary (visual) data control for paper declarations Data entry control and data validation in e-survey system Data validation in ISDMS Data validation module Data validation in Access, preparing data set SQL data base
Primary (visual) data control
Data control in e-survey system
Intrastat data entry control
Data validation in ISDMS Most often mistakes referred to incorrect classifications, incomplete information, methodological problems: not valid CN commodity codes, volumes expressed in supplementary units are missing, volumes expressed in net mass (kg) are missing, ISDMS program identifies: absolute errors, which should be corrected obligatory, possible errors, which could be ignored, if the case is checked and approved as correct.
Data validation in ISDMS Additional checks of the Combined Nomenclature codes, used for seasonal goods (for example, new potatoes in dispatches in wrong period). Credibility checks of some CN codes, goods which are not usual for trade in our country, for example, meat of domestic pigeons, groundnuts seed, tropical wood and so on. Credibility lines - the description of the CN code determines in some cases the net mass of the commodity and/or supplementary unit, for example live swine weighing less than 50 kg, aircrafts of an unladen weight not exceeding 2000 kg. Impossible combinations between mode of transport and CN codes, CN code and country code, for example peat could not be transported by post or only certain goods could be transported by fixed transport installations (pipelines).
Data validation in ISDMS
Price analysis on commodity level 4 8 Price analysis on commodity level (min, max and average prices on based on historical data) and determination of possible ranges for this variation. Min price Average price Max price α% β% β% α% Flag 4 3 2 1 2 3 4 α%=10% β%=80%
Price analysis on commodity level Calculated price Flag Min price Average price Max price
Intrastat data quality control During data validation process statisticians compare Intrastat data with VAT data at enterprise level. If the difference is high, the statisticians call to the enterprise and clarify the reason of discrepancies.
Analysis of aggregated Intrastat data
Number of verified Intrastat rows in 2016 Period Arrivals Dispatches Total Jan 126 789 39 951 166 740 Feb 144 545 43 315 187 860 Mar 149 753 45 377 195 130 Apr 151 939 45 690 197 629 Mai 148 321 46 729 195 050 Jun 147 388 44 613 192 001 Jul 145 688 44 186 189 874 Aug 154 203 46 903 201 106 Sep 156 597 48 262 204 859 Oct 153 461 47 630 201 091 Nov 155 927 48 330 204 257 Dec 147 551 44 009 191 560 1 782 162 544 995 2 327 157
THANKS FOR YOUR ATTENTION!