Methodological note
The TI Global Corruption Barometer 2004 is a worldwide public opinion survey conducted for TI by Gallup International with 52,682 respondents. The Global Corruption Barometer is a set of 5 questions included in the Voice of the People survey 2004, conducted in more than 60 countries by Gallup International members or partners. The TI Global Corruption Barometer is meant to be conducted annually.
Coverage
Overall, the Voice of the People survey was conducted in 64 countries, but some data were missing from individual countries because the authorities did not give permission to conduct certain questions. In Egypt, only question 5 was asked, and in Vietnam, only question 1. In Afghanistan, only questions 1, 2 and 4 were asked.
Timing of fieldwork
The fieldwork for the survey was conducted between July and September 2004.
Demographic variables
The demographic variables, Age, Education, Household income, Education, Employment, and Religion were recoded from their original form in the survey by Gallup International.
Sampling
The sample type is mostly national, but in some countries it is urban only. In most of the countries the sampling method is based on quota sampling, using sex/age/socio-economic condition/regional/urban balances as variables. In some countries random sampling has been done.
The interviews were conducted either on face to face or telephone (mostly in developed countries) with male and female aged 15+ (This information is provided by country in annex 2 of the report on the Global Corruption Barometer 2004).
Weighting
Sample imbalances in the data within a country (e.g. slight corrections to the proportions of age groups, sex, etc.) have been weighted first in order to provide with a representative sample of the national population (or a representative sample of the stated universe, if this is not a total population sample). Subsequently, each country has been weighted to its relevant population (universe). For example, countries where only the urban population was interviewed were weighted up to a total urban population.
Data coding, quality check and analysis
The data coding and quality check has been done by Gallup International, as well as preliminary analysis. The full report of the Global Corruption Barometer has been completed by the Department of Policy and Research at the International Secretariat of TI.
A standard margin of error for the survey is +/- 4.
home
print this page