Validation process and the IT tools used at KAS Kosovo Agency of Statistics Validation process and the IT tools used at KAS IPA training course on Data Validation in the ESS Burim Limolli, Head of Information Systems Mentor Shala, IT expert Luxembourg, May 2017
DATA PROCESSING PHASES Data processing phases after KAS recuperate the completed materials from the field, further activities consiste in codification, data-capture, data editing and imputations, tabulations, analyzes and the preparations for final data dissemination.
DATA PROVIDERS Microdata Aggregated data Surveys on the field Population and Agriculture Censuses Different Administrative data provided by: Microdata Agency for Civil Registration Ministry of Education Business Register Agency Customs of Kosovo Health Ministry Aggregated data Ministries Other public institutions Tax Administration Central Bank of Kosovo Ministry of Finance
DATA COLLECTION METHODS CAPI KAS DATABASE PAPI CAWI PAPI Census Population Agriculture Census HBS CAPI LFS SILC (android tablets) CAWI Tourism statistics Administrative data Based on web services and databases replications CAWI- Hotel Statistics-With this method a link is sent to the respondents via email. They just follow the link to complete the questionnaire independently. CAPI- LFS and SILC-Using a laptop, and tablet , the interviewer can carry on the interview and send back the answers. Data are sent to the main server of KAS. Pacp PAPI- CENSUSES- Personal interviews based on traditonal way (door to door) where the interviewer inputs data on paper questionare
SOFTWARE SOLUTION Data entry Visual Studio & SQL Server SPSS SAS Access LimeSurvey Cspro Validation SPSS STATA Production SPSS SAS Access SQL Server Dissemination and visualation PcWeb StatPlanet
PROCESS OF VALIDATION KAS has no good picture what validations rules are used, but simply it is based on production unit that they do it manually. Usually the data quality is evaluated by different means and methods (PES, data validation, response analyses). The key quality indicators indicate that the data quality can be consider as sufficiently high and acceptable by most of the standard criteria. Those measurement errors that are identified as inconsistencies during the data validation are corrected in the data editing process. Small amount of outlying values are verified by recontacting the respondents by telephone and individually corrected, if needed. However, most of the errors are corrected by using systematic, automatic corrections. For identification and correction of the errors the custom made application, based on a ‘Meta data driven’ approach, was used. Measures used to decrease the measurement errors in most of surveys include many activities with the aim to minimize the measurement errors. The main activities in this respect are: Testing of all instruments Training of field staff, Monitoring of field staff Training of data entry staff. Monitoring of data entry
Data transmission to Eurostat eDAMIS Different format for the data transmission .txt .csv .xml .xls
SDMX SDMX converter of Eurostat SDMX Central Converter of IMF To convert Nacional Account from .xls to xml SDMX Central Converter of IMF Number of Population by sex, Annual Data External Trade Nacional Account CPI PPI LFS
Dissemination Dissemination format Data accessible on the website as well as publications are available free of charges to the public. Some publications are whether in paper-format whether in pdf format also on-line access is ensured through the ASKDATA based on PC-Axis system accessible from the KAS website http://ask.rks-gov.net
Do you have any questions? THANK YOU FOR YOUR ATTENTION!