Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP...

17
Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013

Transcript of Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP...

Page 1: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

Research data workflow

Practice in Slovenian Social Science Data Archives

SERSCIDA WP4 – WORKSHOP Ljubljana

September 2013

Page 2: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

SIP, AIP, DIP

• Submission Information Package (SIP)• Archival Information Package (AIP)• Dissemination Information Package (DIP)

SIP DIP

AIP AIPLong term preservation

Page 3: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Recommended formats – input

Type of material Recommended formatOther acceptable

formats

Questionnaire

Rich Text Format (*.rtf) structured metadata record of

questionnaire (*.xml) by DDI or CAI programme (*.bmi)

other text formats (*.docx, *.txt, etc.)

*.pdf or other graphical formats

printed version

Data material(data file)

SPSS (*.por, *.sav) plain text data, ASCII (*.txt) +

structured text or mark-up file containing metadata information (variable names, labels, categories, question text)

other statistical packages

tables (*.xlsx etc.) data bases

Textual material (study description, codebook, interviewer instructions, speech to respondents, copies of research reports)

Rich Text Format (*.rtf) printed version *.pdf or other

graphical formats other text formats

(*.docx, *.txt, etc.)

Page 4: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Recommended formats – distribution

• STUDY DESCIPTION: DDI structured XML• DATA FILE: ASCII + xml distributed in

formats that can be exported from Nesstar

• OTHER TEXTUAL MATERIAL: PDF

Page 5: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Recommended formats – archiving

• DATA FILE:ASCII (*.txt) + xml with DDI file and data description

Page 6: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Recommended formats – archiving

• QUESTIONNAIRE, TEXT MATERIAL:original (any format) + distribution files (PDF)

• STUDY DESCRIPTION:DDI structured XML

Page 7: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Licence AgreementFree:• to Share — to copy, distribute and transmit the work• to Remix — to adapt the work• to make commercial use of the work

Under the following conditions:Attribution — You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work).

Free:• to Share — to copy, distribute and transmit the work• to Remix — to adapt the work

Under the following conditions:Attribution — You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work).Noncommercial — You may not use this work for commercial purposes.

Page 8: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Naming files and versioning

File format:StudyID_MaterialType_Language_Version_Subversion.FileFormat

Example: sutr1006_p1_sl_v1_r2.txt

URN:URN:SI:UNI-LJ-FDV:ADP:StudyID_MaterialType_Language_Version

Example: URN:SI:UNI-LJ-FDV:ADP:sutr1006_p1_sl_v1

Page 9: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Managing workflow

• Project tracking software

• Task for every study, with 29 subtasks covering:- general part with email correspondence- managing deposited materials- preparing data file- preparing study description- publishing

http://nesstar2.adp.fdv.uni-lj.si:8080/browse/RAZ-4536

Page 10: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Cleaning data and documentation

• Frequencies check• Variable names, values• Missing values• Recode• Weight• Anonymisation • Cumulative dataset

Page 11: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

Anonymisation

Sebastian Kočar

Expert Assistant in Social Science Data Archives

SERSCIDA WP4 – WORKSHOP Ljubljana

September 2013

Page 12: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Anonymisation in the archives - types

• basic anonymisation - of mostly academic research dataset

• anonymisation of Eurostat files

• anonymisation of official statistics Public Use Files (PUF)

Page 13: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Basic anonymisation of distributed microdata in archives

• deleting variablesDirect identifiers (telephone numbers, addresses etc.) are removed.

• recoding indirect identifiers But still allowing serious researchers to receive datasets with indirect identifiers non-recoded). Recoding includes removing values and bracketing – combining the categories of a variable.

Page 14: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Anonymisation of Eurostat files (the case of Eurostat Labor Force Survey)

• deleting variables: indirect identifiers and unneeded variables are removed (municipality, wave nr. etc.)

• bracketing: age, marital status, education, years of residence, age of establishment of residence, duration of search of employment, professional status, country & nationality

• classification: income numbers are not given, respondents are divided into classes based on their income

• aggregation: economic activity and occupation values are aggregated at 1-digit level

• top-coding: restricting the upper range of a variable (nr. of hours worked)

Page 15: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Anonymisation of official statistics Public Use Files for distribution in archives

• anonymisation software: μArgus, R! (sdcMicro, bethel, sampling packages), Cornell anonymisation toolkit, synthetic data generators

• anonymisation technics: data reduction techniques (global coding, local suppression etc.), data perturbation techniques (micro-aggregation, PRAM etc.), sampling, generating synthetic microdata

Page 16: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu

Anonymisation – a case study

• PUF prepared in cooperation with SORS Sector for General Methodology and Standards

• anonymisation procedure which follows Eurostat LFS anonymisation criteria (in SPSS)

• calculating individual and global risk (R! – sdcMicro)

• calculating strata allocation, based on individual risk averages by strata (R! – bethel)

• stratified sampling, based on the inclusion probability of a certain case (R! – sampling – samplecube)

• sample weights recalculation

• LFS 2010 PUF distributed in August 2013

Page 17: Research data workflow Practice in Slovenian Social Science Data Archives SERSCIDA WP4 – WORKSHOP Ljubljana September 2013.

www.serscida.eu