Transcript MICS Data Processing Workshop
Multiple Indicator Cluster Surveys Data Processing Workshop Overview of Data Processing System
MICS Data Processing Workshop
Content of the Presentation
• Overview of the MICS data processing system • Data processing using paper questionnaires • Main characteristic of the MICS CAPI system • Creating analysis files • Data archiving
Content of the Presentation
• Overview of the MICS data processing system
MICS Data Processing System: Actors and Roles – Country data processing manager and country team: • Customization of data entry programs, data entry, editing, and production of datasets • Customization of tabulation syntaxes and tabulation – Regional Office MICS Coordinator • Coordination and supervision, organization of the Data Processing workshop – Regional Office Data Processing Consultant • Technical support and review of customized programs and close work with country teams – HQ Data processing unit • Development of standard programs, templates and coordination of Data Processing workshops
MICS Data Processing System
• The data-processing system can be divided into following phases: – Customization of MICS data entry/collection program and tabulation syntaxes, – Establishing the data entry system locally, – Primary data processing (data entry/data collection), – Secondary data processing (creating analysis files), and – Tabulation – Archiving
Content of the Presentation
• Overview of the MICS data processing system • Data processing using paper questionnaires
MICS Data Processing System
• Designed to deliver the first results of a survey within several weeks after the end of fieldwork • Such rapid turnaround time is possible when completed questionnaires are entered simultaneously with survey fieldwork • Data for each cluster is stored in a separate data file and is processed as soon as all the questionnaires from a cluster are returned from the field • This approach breaks data processing down into discrete segments and allows it to progress while fieldwork is ongoing • By the time the last questionnaires are finished and returned to headquarters, most of the data have already been processed
Primary Data Processing Flow
Main Data Entry Structure Check Verification Data Entry Difference Listing Backup Raw Data Secondary Editing Backup Final Data
Primary Data Processing
• Main data entry – First time data is entered • Structure check – Checks structure of data files • Verification data entry – Second time data is entered • Difference listing – Two data files are compared; differences resolved
Primary Data Processing
• Raw data backup – Verified data are backed up to a directory containing raw data files • Secondary editing – Complex inconsistencies are investigated and resolved if able • Final data backup – Edited data are backed up to a directory containing edited data files
Data Processing Personnel
• Questionnaire administrators • Data entry operators • Secondary editors • Data processing supervisor
Questionnaire Administrators
• Receive clusters from the field • Check that all questionnaires are present • Check that questionnaires are ready to enter • Check that HH and individual questionnaires are in the proper order, and rearranges them if not • Keep track of location of all clusters • Should follow interviewer training
Data Entry Operators
• Enter main data • Enter verification data • Resolve differences between files • Must follow interviewer training • Must be familiar with the questionnaires
Secondary Editors
• Investigate complex inconsistencies • Tell supervisor if and how to resolve inconsistencies • Review editing guidelines • Must be present during interviewer training • Need excellent understanding of questionnaire
and goals of survey
Data Processing Supervisor
• Resolves data entry problems • Maintains programs • Oversees entire data processing system • Must be present during interviewer training • Must have excellent grasp of questionnaire • Must have programming skills in SPSS and CSPro
Questionnaire Administrator Training
• Review list of checks in data processing chapter • Give QA several clusters and check work • Establish questionnaire storage procedures
Data Entry Training
• Begin when you have one cluster for each data entry operator • Allows you to – Train data entry operators – Debug programs • Practice verification at the same time • When you have finished – Fix entry programs – Delete data files
Secondary Editor Training
• Wait until you have 3-4 clusters double-entered • Give secondary editors – Copy of editing guidelines – A cluster’s error listing – The cluster’s questionnaires • Review work with secondary editor • Try to schedule a day when data entry operators aren’t working
Data Processing Equipment
• Data entry machines – Windows XP, 2000, Vista,7 or 8 – Supervisor’s machine – Windows XP, 2000, Vista,7 or 8 – Available disk drive space for all software and MICS data files • Uninterrupted power supplies (UPS) • Network • Surge Protectors
Data Processing Equipment
• A printer • Paper • Toner cartridges/printer ribbons • Flash drives (if network is not established) • Green pens
Data Processing Rooms
• Data Entry – Desk space for the monitor, keyboard, and stack of keyed and unkeyed questionnaires – Area to store “in-progress” clusters • Editing – Quiet space for editors to work • Questionnaires – Must contain means (e.g., shelving) to organize questionnaires by cluster
Server and Keyer’s Directory Structure MICS5 CSPRO DATA ENTRY
contains data entry programs
VERI
contains data from main entry contains data from verification entry
Supervisor’s Directory Structure
MICS5 CSPRO
BACKUP DICTS ENTRY EXPORT FINAL GPS RAW SUPER
SPSS WEIGHTS
backup of verified data dictionaries used in all programs (not just entry) contains data entry programs programs to transfer data from CSPro to SPSS a copy of all edited data GPS data entry program (if applicable) contains all unedited data (just after differences resolved)
all programs not related to data entry phase
contains all SPSS programs used for export and tabulation contains weights spreadsheet and data files
Data Entry
Cluster reaches the office Cluster acceptance is done. All the questionnaires are counted and recorded according to their type and result code to the cluster tracking form (CTF).
Cluster is sent to Supervisor Supervisor records the information on CTF to the system. Cluster is ready for main data entry. Supervisor gives it to first keyer and records keyers information to system and to CTF. Supervisor verifies the main and verification data by the supervisor menu. If there is typing errors, lets the keyers correct them at their computers. If no problem, the raw data is backed up. Sytem automatically gives date, it should be written to CTF MAIN DATA ENTRY First keyer enters the whole of the questionnaires of the cluster.
VERIFICATION Second keyer enters the whole of the questionnaires of the cluster.
Supervisor runs the structure check control. If all questionnaires are entered, gives to second keyer for verification and records this information to system and CTF.
After backing up raw data
After backing up of raw data: Secondary editing is done by the supervisor menu. NO ERROR “Back up the final data” ERRORS Secondary Editor: By using the “Editing Manual “ the necessary changes at the data with “Modify the data” option at Supervisor menu is done. Supervisor follows the same procedures for every cluster If all procedures are completed for all clusters: Export the data to SPSS
NETWORK STRUCTURE
Network Computer: \\MICS5\CSPRO\DATA \VERI \ENTRY\entry.enc
\entry_menu.enc
Keyer Computer: \\MICS5\CSPRO \DATA \VERI \ENTRY\entry.enc
\entry_menu.enc
\entry_menu.pff
Supervisor Computer: \\MICS5\SPSS \\MICS5\CSPRO\BACKUP \DICTS \ENTRY \EXPORT \FINAL \GPS \RAW \SUPER
Content of the Presentation
• Overview of the MICS data processing system • Data processing using paper questionnaires • Main characteristic of the MICS CAPI system
Mobile data collection
• In the recent years we have seen development of many innovative data collection tools using handheld PocketPC personal digital assistants (PDAs) , smartphones or tablet personal computers .
Tablet Personal Computer
• Tablet – sized computer with the key features of a full – size personal computer • Various operating systems • Computer vs. cell phone capability
Personal Digital Assistants
• Handheld computers • Various operating systems – Windows mobile – Palm – Others (e.g. iPhone, Nokia) • Cell phone capability – Plus: good communications – Minus: security
Tablet vs PDA
• Choosing a tablet over a PDA - Tablets have much larger screens, better resolutions, more space for on screen typing making things easier to see and more useful for data entry - Tablets and PDAs have similar battery life, though if PDAs are used for making and receiving calls battery will need more frequent charging - Better security
Hardware requirements for CSPro
Tablets:
• Required configuration: Full Microsoft Windows 7 or 8 (NO windows RT tablets)
PDAs:
• Required configuration: Windows Mobile versions 5 and 6 (note that UNICODE is not supported)
MICS CAPI System
• All applications to collect and administer data are written in CSPro • Three systems: – Interviewers - data collection – Tablet/PDA – Supervisors – data monitoring and control – Tablet/PDA – Central Office - centralized data and monitor fieldwork - PC
Repair Utility
Int 2 Int 5
Interviewer’s System Supervisor
System Updates
Update Utility
Screen Reports Reports
Supervisor
System Updates
Central Office Supervisor’s System Central Office System
Training
• 2 weeks training (paper questionnaires) including 3 days field exercise + 1 week extra training (PDAs/Tablets) + 4-5 days extra field exercise • High Level Technical Support
Directories on Interviewer's Tablet
\My Documents\MICS5\
• Data Data from the questionnaires as collected during the interview • Dicts Dictionaries describing the data structures • Entry Questionnaire and Interviewer’s menu programs • Ref Reference data files used by the programs • Utility Utility programs used for Bluetooth transfer, compressing files,
upgrading programs, etc.
• Work Working files used by the menu and programs
\SD Card\
• Backup Backup copy of the data files from Data created after the
interviews
Directories on Supervisor’s Tablet
\My Documents\MICS5\
• Receive • Temp • Work • Reports
Directory for data received from the interviewers Temporary directory for working files transferred from interviewers, prior to transferring to the Receive directory Working data files Report files created by supervisor applications
Content of the Presentation
• Overview of the MICS data processing system • Data processing using paper questionnaires • Main characteristic of the MICS CAPI system • Creating analysis files
Secondary Data Processing
• Exporting data from CSPro – Create SPSS data file and syntax file from CSPRO data file and dictionary • Importing data into SPSS – Executing syntax file created by CSPro • Recoding variables – Creating new variables and recoding old variables
Secondary Data Processing
• Adding sample weights – Sample weights are added from weights spreadsheet • Adding wealth index – Wealth index calculated then added to files • Adding GPS data – Geographic location data added to files • Tabulation – Tables are generated from the analysis files
Content of the Presentation
• Overview of the MICS data processing system • Data processing using paper questionnaires • Main characteristic of the MICS CAPI system • Creating analysis files • Data archiving
Data archiving: rationale
• Collecting data is expensive. Data should be used beyond producing basic report.
• Survey microdata are valuable resources for government departments and academic researchers. • Survey data constitute valuable and irreplaceable assets which should be managed in a way that encourages their widest possible use and re-use. • At the same time, data collectors main focus should be protecting respondents while making microdata assessable.