Managing ICT resources
Download
Report
Transcript Managing ICT resources
Managing ICT resources
ICT - a core management issue
Part 1
Produced in Collaboration between World Bank Institute and
the Development Data Group (DECDG)
Copyright 2010, The World Bank Group. All Rights Reserved.
Impacts of ICT
• ICT developments are dynamic
• NSOs have been particularly affected
• Five key developments:
1. Powerful, relatively inexpensive computer equipment is
available to most staff members in many statistical offices
2. User-friendly applications software gives control over a number
of key statistical functions
3. Readily available off-the-shelf components have made inhouse application programming easier
4. Computer networking facilitates internal access to data and
metadata
5. Information technology gives access to external sources of
information
2
Copyright 2010, The World Bank Group. All Rights Reserved.
Models of ICT management
There are two broad models of managing ICT in NSOs
1. Model 1 is based on strict standardization and centralization of
decision-making
2. Model 2 is more loosely structured
• In practice, most models are a hybrid of models 1 and 2
3
Copyright 2010, The World Bank Group. All Rights Reserved.
Model 1 vs. Model 2
•
Both model 1 and model 2 have advantages
•
Under a strict model 1 regime, personal software and data import and
export on individual machines are prohibited
•
This has security advantages with respect to both viruses and confidential
data
•
Under both models, hardware procurement is probably centralized
•
The general understanding today is that software applications should be
developed as closely as possible to the substantive user
•
This can be achieved under both models
Copyright 2010, The World Bank Group. All Rights Reserved.
4
Old and new concerns
•
In the mainframe period recommendations were:
• Human IT resources are scarce and must be kept under central
supervision
• Document all systems extensively
• Organize the workforce balancing functional IT and subject matter
experience
•
Modern recommendations include:
• Do in principle not subcontract IT development work, other than certain
one-time applications, to outside agencies
• Do not assign IT staff to new projects until they have completed clear
documentation of data and programs for ongoing projects
• Do not leave administration of hardware to other parties unless it is
guaranteed that statistical processing will get the highest priority.
5
Copyright 2010, The World Bank Group. All Rights Reserved.
Use of ICT in statistics
• Modern NSOs use ICT throughout all operations:
•
•
•
•
•
Data capture
Data processing
Data analysis
Data tabulation
Data dissemination
• Key issues include:
• Using standard products or custom-made software
• How and when to develop custom-made software
6
Copyright 2010, The World Bank Group. All Rights Reserved.
Statistics production model
• Production model is a cornerstone for ICT policy
• Model consists of statistical registers input data (microdata)
• Data aggregated from input data are stored in an output database
with very detailed aggregations
• Output database is available to all staff for tabulations and
answering inquiries
• Publication database is a more aggregated version of the output
database and is accessible for the public through the NSO website
• A common metadata base supports the entire database system
7
Copyright 2010, The World Bank Group. All Rights Reserved.
ICT for data capture
• Data capture can be ‘manual’ or computer-assisted
• Another option is collecting data from registers
• In addition to traditional paper methods, many NSOs focus on
electronic data collection
• Free or cheap software for entering, editing, and tabulating of survey
data includes CSPro and Blaise
• CSPro is available from the United States Census Bureau
• Blaise can be obtained from Statistics Netherlands
• Both packages are used worldwide
8
Copyright 2010, The World Bank Group. All Rights Reserved.
ICT for data processing
• Spreadsheets are only useful for very simple processing
• Blaise and CSPro software can handle more complex projects
• Blaise data entry program can store data in Oracle or SQL server
databases
• Blaise/Manipula application can be used for auto-coding, checking
plausibility and validity or to clean data
• Blaise can generate a log file with time measurements to track down
what happened during an interview
9
Copyright 2010, The World Bank Group. All Rights Reserved.
ICT for analysis
• CSPro and Blaise also offer analysis tools
• Comprehensive commercial software for statistical analysis includes
SPSS and SAS
• SAS has more functionalities than SPSS, but is also more
expensive
• For most NSOs SPSS is powerful enough
10
Copyright 2010, The World Bank Group. All Rights Reserved.