editing of linked micro files for statistics and research
TRANSCRIPT
Editing of linked micro files for statistics and research
Data sources Background information for editing Creating linked files Employee jobs Confidentiality and privacy
Datasets and input
Administrative Data sets
Base RegistersStatistical Surveys
EnterprisesInternal DataSystems
Web-questionnaire
PaperQuestionnaire
Editing at response time
EnterprisesInternal DataSystem
Web-questionnaire
Checks and corrections interacting
with respondent
Collecting information from the prosesses
Editing using prosessinformation
Statistical Surveys
PaperQuestionnaire
Collected information from the prosesses on web-questionnaires
Editing withoutinteraction with
respondent
DynarevDynamic revision in business surveys
Data collection
Publishing
Totalpopulation
Sampled population
MetadataEnterprise accounts,other administrative data
and statistical surveys
Dynarev
Data collection
Publishing
Totalpopulation
Sampled population
MetadataEnterprise accounts,
other administrative dataand statistical surveys
Calculation-basis and rules
Checkingrules
Editing Calculating
Data sources for research
The strength of administrative data
Longitudinal databases Base registers Education Labour force Income Transfers Annual Census file Health indicators
Infrastructure for creating linked files
The advantage of use of linked files
Base registers and official ID numbers
Consistency control
Creation of linked files - centralisation
Editing and imputation
Longitudinal data Base register of person Base register of business Base register of dwelling Enterprise accounts Labour Force
Employee jobs by source Norway
Two sources A = Tax Agency B = Social Security
Jobs in A and B 1 911 000 Jobs only in A 169 000 Jobs only in B 26 000 Total employee jobs 2 106 000
Editing and imputation
Annual Census micro file
Activities and source of livelihood
Micro file on generations
Editing and imputation of linked files for administrative purposes
E-government – a variable should be reported only once
A new agency to reduce number of persons not in the labour force
2 400 000 in Labour Force 700 000 not in LF
Editing and imputation by documented corrections
Confidentiality and privacy
Administrative data are often sensitive Data Inspectorate Classification Identifiable – ID number De-identified – ID number removed Anonymised – not possible to link to
other sources