Data Collection Optimization

Transcription

Data Collection Optimizationfirst attemptÁgnes AndicsBusiness Statistics, Methodology Department, HCSOagnes.andics@ksh.hu14.09.2015Statistical Data Editing Work Session, Budapest

Question2

Cumulative percentage of Questionnairearrivals of survey OSAP 17993

Reminder system§ 1st reminder email for existing email addresses, automatic 2 days before deadline§ 2nd reminder email 1 day after the deadline, automatic§ 3rd reminder email/letter suggested 3 days after the deadline§ 4th reminder mail Snail mail with recorded delivery (only after the 3rdreminder)§ Phone calls4

Daily sum of the “total product in naturalunit” in survey OSAP 10395

Key elements Key variables Key respondents with high impact factor Response rate Rate of over-coverage6

Pre-Estimation Procedure Supportedby IndicatorsGoal: create a highly automated procedure that enables the HCSO to track the progress of theestimations made at any moment of time during the data collectionperiod and indicates whether the available data for a givensurvey is good enough both in quantity and quality.7

Steps of the procedure EditingØ selective – for key respondentsØ automatic – for key variables ImputationØ item – for key variablesØ unit – for key non-respondents Pre-estimation – for key variablesv quality indicatorsv significance level?8

Steps of implementation Getting to know each survey process Analyse the past based on meta information More automatic editing and selective editing Imputation policy for each survey process Appropriate accuracy measures with significance level What else?Thank you!9

IT systems in HCSOBusinessregisterAgricultural, Social, Nonprofit organizations, Spas,Movies,Commercial Accommodations etc.Survey ManagementSystemGÉSASurveyspaper questionnairesSurveysSpecial registersE-questionnairesExcel questionnairesElectronic DataCollectionELEKTRAData transmissionsecondary dataData CollectionKARÁTData ValidationADELControl systemADAMESDataProcessingEARMETADataWarehouse10

Data Collection Optimization first attempt Ágnes Andics Business Statistics, Methodology Department, HCSO agnes.andics@ksh.hu 14.09.2015 Statistical Data Editing Work Session, Budapest . Question 2 . . ADAMES Data Warehouse . Created Date: 9/14/2015 6:09:51 AM .