Abstract

Surveys and polls, considered as both research techniques and social phenomena, are a central part of everyday life in contemporary society. From a methodological standpoint, surveys can usually be identified by the following elements: the individual as collection unit; recruitment and training of interviewers; administration of a structured questionnaire; construction of a data matrix; recourse to statistical data analysis. Sources of survey finding variability and problems with their generalisation to wider populations (especially in the case of opinion polls) are examined. Particular attention is paid to the effects of recent technological innovations in data collection techniques (telephone- and computer-assisted interviewing). As a social phenomenon, surveying represents a key activity for the economy (market research, international survey institutes, professional associations), politics (monitoring of public opinion), and the mass media (agenda-setting functions, newsmaking through "pseudo-events"). The convergence of economic, political, and media interests has increased the potential for consensus manipulation in democracies; along with other developments, this has led some groups to call for more stringent regulations on survey activity and publication of poll findings.