“Public Goods through Private Eyes” Project. General information
“Public Goods through Private Eyes” is a full-scale comparative public opinion survey carried out in 2013 and 2014 as a result of the 2009 ERC funding of 1.73 million Euro awarded to dr hab. Natalia Letki (project PI, University of Warsaw). The aim of the project was to collect high-quality survey data on attitudes and behaviour towards public goods and the state in 14 post-communist countries:
Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Moldova, Poland, Romania, Serbia, Slovakia, Slovenia and Ukraine.
PGPE Questionnaire
The questionnaire was developed by the PGPE team with support from the Project’s Advisory Board: Rene Bekkers (VU University Amsterdam), Klarita Gërxhani (EUI), Erich Kirchler (University of Vienna), Stephan Muehlbacher (University of Vienna), Kristina Murphy (Griffith University), Pamela Paxton (The University of Texas at Austin) and Michael Wenzel (Flinders University). It contained 7 main modules:
A. Local community
B. Social trust and social cohesion
C. Personality and commitment
D. Public goods
E. Institutional quality
F. Tax behaviour and law compliance
G. Green behaviour
H. Political participation
I. Socio-economic background.
The survey also contains 4 vignettes – fully randomized survey-embedded experiments.
In each country respondents were clustered in 75 stratified sampling points (with the exception of Ukraine, where Russian annexation of Crimea interfered with fieldwork and only data in 74 sampling points were collected). The sampling points were designed to cover a spatially relatable area that respondents could refer to when contextualizing their survey responses. For details, see PGPE document on sampling procedure.
Result
The result of the project is a multi-level multi-component survey, comprising of:
1. The main survey of 22039 respondents nested in 1049 SPs in 14 countries.
2. The survey of interviewers (each interviewer had to complete the survey prior to commencing work for the project).
3. Survey of ecological data for the PSU level in all countries.
Surveys 2 and 3 can be linked to survey 1 to explore interviewer effects (survey 2) and contextual effects (survey 3). For confidentiality reasons, surveys 2 and 3 cannot be released.
Sampling procedure was developed in cooperation with dr Matthias Ganninger, dr Sabine Haeder and dr Siegfried Gabler (GESIS Mannheim) (for details, see a document on PGPE sampling).
Weights were prepared based on the ESS standards by Jan-Philipp Kolb (GESIS Mannheim). The following weight types are provided in the PGPE dataset: DWEIGHT (design weight), PSPWGHT (post-stratification weight), PWEIGHT (population weights) and PPSPWGHT (combination of post-stratication weights and population size weights).
Main fieldwork was carried out by two companies and their subcontractors: in Poland it was carried out by CBOS, in all other countries - by IPSOS Strategic Marketing. PAPI and CAPI were used as data collection methods.
Quality
To achieve as high a quality as possible the key following principles were applied:
· Sampling was designed and carried out centrally, by the PGPE team in cooperation with sampling experts.
· Pre-listing of address was verified centrally, by the PGPE team, on the basis of maps.
· Questionnaire was translated following the ESS round 5 guidelines.
· Questionnaire was tested qualitatively, and a quantitative pre-test on quota samples were carried in each country prior to the main survey.
· Data was centrally checked and screened for inconsistencies and feedback was given to the fieldwork companies to be taken into account in the subsequent phases of fieldwork.
· Post-survey control in most countries was carried out with the participation of the PGPE team members.
Translation of the questionnaire
Translation guidelines were modeled after ESS round 5, with particular emphasis on keeping the equivalence of meaning and the symmetry and consistency of scales. Translation was carried out in three stages: i) questionnaire was translated by two independently working, experienced translators per language; ii) two versions were adjudicated by an experienced adjudicator; iii) language versions used as minority questionnaires in other countries were harmonized by country coordinators
Questionnaire testing
The PGPE Questionnaire was tested in two stages. First, a Qualitative pre-test was performed in Polish by GFK Polonia in 3 waves of question testing in a studio with a venetian mirror. Second, a Quantitative pilot was performed by fieldwork companies on a representative quota sample of N=80 in each country. The aims of both qualitative and quantitative pre-testing was to control the quality of translation and to shorten the questionnaire.
(2025-02-25)
