d ata p rocessing w orkshop bangkok, thailand, 15-19, sept 2008 by mr. pen socheat, nis, cambodia 1
TRANSCRIPT
DATA PROCESSING DESIGN
Data processing is organized around EA batch
There is one set of data files for each EA Allows us to
Process data in parallel with data collection Allows feedback to the field Process data in discrete segments Keeps size of data files manageable
Data file names include geographical information code
2
DATA PROCESSING DESIGN (CONT.)
Data processing is split into two phases Primary Secondary
Goal of primary phase Clean, edited data
Goal of secondary phase tables and Analysis files
3
PRIMARY DATA PROCESSING FLOW
Main Data Entry
Structure Check
Verification Data Entry
Verification
Backup Raw Data
Secondary Editing
Backup Final Data4
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 Verification
Two data files are compared; differences resolved
5
PRIMARY DATA PROCESSING
Raw data backup Verified data are backed up to a separate
directory Secondary editing
Complex inconsistencies are investigated Final data backup
Edited data are backed up to a separate directory
6
CONTROL SHEET Keeps track of data processing One row for each person Enter
Dates each task completed Number of data entry operators
7
SECONDARY DATA PROCESSING FLOW
Export Data from CSPRO
Import Data into SPSS, STATA
Recode Variables
Add GPS Data
Run Tables
8
SECONDARY DATA PROCESSING
Exporting data from CSPRO Create SPSS data file and syntax file from CSPRO
data file and dictionary Importing data to SPSS, STATA
Executing syntax file created by CSPRO Recoding variables
Creating new variables and recoding old variables
9
SECONDARY DATA PROCESSING
Adding GPS data Geographic location data added to files
Tabulation Tables are generated from the analysis files
10
DATA PROCESSING PERSONNEL
Questionnaire administrators (Logistic) Data entry operators Secondary editors Data processing supervisor
11
QUESTIONNAIRE ADMINISTRATORS
Receive questionnaire from the field Scan questionnaire barcode represent
Geographical data from the field Check that all questionnaires are present Check that questionnaires are ready to store Should follow the instruction of questionnaire
administrator
12
DATA ENTRY OPERATORS
Enter main data Enter verification data Resolve differences between files
Follow the instruction manual of data operator
13
SECONDARY EDITORS
Investigate complex inconsistencies Tell supervisor if and how to resolve
inconsistencies Review editing guidelines
14
DATA PROCESSING SUPERVISOR
Resolves data entry problems Maintains programs Oversees entire data processing system
Must have excellent grasp of questionnaire Must have programming skills in SPSS and
CSPRO
15
DATA ENTRY TRAINING
One week for training Train data entry operators Debug programs
Practice verification at the same time A few day practice
When you have finished Fix entry programs Delete data files
16
DATA PROCESSING EQUIPMENT
Data entry machines Intel(R) Core(TM) 2 Duo, WinXP professional+,
1.6 Gb RAM, 100 Gb hard drive & DVD/RW rewritable CDROM
Supervisor’s machine Intel Core 2 Duo, WinXP professional 1.6 Gb RAM,
100 Gb hard drive & DVD/RW rewritable CDROM, secondary storage device (USB 1.0/2.0 GB)
Uninterrupted power supplies
17
DATA PROCESSING ROOMS
Data Entry Room for computer
Editing Quiet space for editors to work
Coding Quiet space for coding to work
19
DATA ENTRY DIRECTORIES STRUCTURE
Data Main data entry files
Dicts CSPRO dictionary
Entry Data entry programs
Veri Verification data entry files
21
SUPERVISOR CSPRO DIRECTORIES STRUCTURE
Backup Backup of verified data
GPS (if applicable) GPS data entry program
Export Programs to transfer data
Final Backup of edited data
Raw Data from data entry machines
22