The public opinion survey conducted by CBOS (Public Opinion Research Center) on a representative nationwide sample of adult Poles (N=1,024) in February 2004 has included a set of questions developed in order to to reconstruct various dmensions of respondents’ attitudes towards public opinion polls (e.g. anonymity, confidentiality, role of opinion polling in a democratic society, uses and abuses of public opinion polls, etc.). The aim was to verify the hypothesis that people who are better educated, more engaged in politics and already have some survey experience tend to define the interview situation in amore accurate and more positive way than those who are less educated, less involved in politics and having no survey experience at all. Thus, education, political participation and survey experience were supposed to be the main factors determining positive attitudes towards the interview, consequently leading to more accurate cognitive processes, lower tendency for satisficing and better quality of answers. Factor analysis (to distinguish patterns of defining public opinion polls) and regression analysis (to verify the hypotheses) were applied in the analysis of findings