Title here - United Nations Statistics Division

International Seminar on Population and Housing Censuses:
Beyond the 2010 Round
Seoul, Republic of Korea, 27-29 November 2012
Use of the internet response
for censuses of the 2010 round
in the UNECE region
Paolo Valente
United Nations Economic Commission for Europe
Statistical Division
Content:
1.
2.
3.
4.
5.
Why offering the internet response
Pioneer countries: experiences in 2000-2006
Internet response in 2010 round in UNECE region
Experiences in selected UNECE countries
Benefits (preliminary information)
1. Why offering the internet response?
Main drivers:
•
•
•
•
•
To maintain/increase census participation (coverage)
To improve data quality
To respond to expectations and demand by the public
To comply with « e-government » policies
To respond to environmental concern: less paper and
transport of forms/enumerators
• To reduce costs in the long-term
No cost reduction in short and medium term
2. Pioneer countries: Experiences and take up rates
in 2000-2006 (table prepared in 2008)
Country
2000-01
2005-6
2010-11 (plans)
Canada
New Zealand
Australia
Rep. of Korea
Small test
----
18%
Increase to 35-40%
Increase to 20-30%
Increase take up
Increase take up
Singapore
England Wales
T.u. unknown
--
7% (low pub)
9%
1% (low pub)
n.a.
n.a.
Cyprus, Hungary, Ireland, Italy, Estonia, Poland…
Spain
Switzerland
United States
Low take up
4.2%
--
n.a.
n.a.
7.2% (test)
Increase take up
Estimated 20%
Under evaluation
?
No (register-based)
Not offered
3. Internet response in 2010 round in UNECE region
Context:
Diversification of census methodologies
- various forms of «combined census» based on data from registers
and field data collection
Frequent adoption of multiple enumeration methods
Internet response adopted in many UNECE countries
as optional method, as part of:
- traditional census
- combined census (registers + field data collection)
3. Internet response in 2010 round in UNECE region
Traditional census:
Internet response offered in 7 countries (out of 31)
- Common in countries using also self-enumeration with paper
questionnaire (6 countries out of 11)
-
Canada
Hungary
Luxembourg
Portugal
Slovakia
United Kingdom
- Rare in countries using also face-to-face interviews (1 out of 20)
-
Bulgaria
3. Internet response in 2010 round in UNECE region
Combined census (registers + field data collection):
Internet response offered in 10 countries (out of 12)
-
Czech Republic
Estonia
Germany
Italy
Latvia
Liechtenstein
Lithuania
Poland
Spain
Switzerland
4. Experiences of UNECE countries (1/2)
Country
Strategy
Take up rate
Estonia
Internet response used as main enumeration method
1) Internet census (one month)
2) Enumeration with paper forms
66%
(target: 25%)
Canada
Wave methodology, with internet as preferred mode:
- Wave 1, gr. 1 (75%): Letters with internet access code
- Wave 1, gr. 2 (25%): Questionnaire and codes mailed
- Waves 2-4: mail, telephone, face-to-face reminders
54%
(target: 40%)
Portugal
Both paper forms and internet ID codes distributed
1) Internet data collection (one week)
2) Internet + paper-based collection (2 weeks)
3) Only paper-based collection
50%
Lithuania
Priority give to internet census
1) Internet census (two weeks)
2) Enumerators visit and interview non-respondents
34%
4. Experiences of UNECE countries (2/2)
Country
Strategy
Take up rate
Italy
Paper questionnaires and internet access codes mailed
Return possible in different forms, including: internet,
enumerators, post offices and census collection centres
33%
(in pilot 9%)
Latvia
- Internet census planned for 10 deys, extended by 2 days 30%
- Field data collection with paper questionnaires
- Additional opportunity for internet response, for 10 days
Czech
Republic
Paper questionnaires and internet codes distributed
Returns by internet or on paper possible (4 weeks)
27%
United
Kingdom
Paper questionnaires and internet codes distributed
Returns by internet or on paper possible
Internet not specifically encouraged
Main objective was maximizing overall response
15%
(target 25%)
5. Benefits of internet response
(preliminary reports)
Canada, UK:
• Better data quality
• Lower item non response
 Thanks to use of filters, controls, warnings
UK:
• Foreign-born respondents more likely to use
internet response
 Possibility to translate questions online  Better quality of responses
Internet response in 2010 round in UNECE region
More information expected from forthcoming UNECE
online survey on national practices in 2010 round:
-
Take up rates
Strategy for internet and paper returns
Methods used to distribute internet access codes
Data control features used
Methods used to ensure data security/privacy
Strategies used to cope with peak demands