Opinion poll
The opinion research (also: demo copy ( ancient Greek δῆμος démos "people", σκοπείν skopein " peek " ) or survey research ) is used to identify opinions, that is, insights, attitudes, moods or desires of the population. It represents the counterpart to Ökoskopie or ökoskopischen market research that determined using empirical methods objective findings ( sales, price Entwicklunge, buyer structures, etc. ).
For the consumer surveys collected primary data through surveys based on a representative cross -section of the population to be examined and then interpreted. The survey can be either in person, by telephone, in writing or through an online questionnaire. In the opinion research can both one-time surveys ( cross-sectional studies ) as well as to long-term studies ( longitudinal studies ) Go.
Long-term studies can be applied either in the form of repeated surveys of different sampling techniques (trend study) or in the form of panel studies ( personal identical multiple survey ) in which the same individuals are repeatedly interviewed over a period of often several years.
History
The first documented opinion survey was conducted in 1824 by the local newspaper in Harrisburg. It was asked who becomes president after the election of 1824. 59 percent stated that Andrew Jackson would be the next president. When choosing this also got the most votes, but not the most electors and so was John Quincy Adams 6th President of the United States of America. In the following years, methods were developed for representative survey, including by George Gallup, who developed the so-called Gallup poll sampling methods.
In 1947, Elisabeth Noelle - Neumann, the first research institute in Germany, the Institute for Allensbach. It emerged soon other polling organizations, in 1955 founded the Working Group for business market and sales research eV.
Problems in the polling
Opinion research purport to reflect the opinions of certain population group; occasionally also public opinion. The branch of empirical social research always leads to original empirical data. However, the quality and validity of the empirically determined ( mathematically correct ) survey results depends on the quality and objectivity of their methodological system. Without knowledge of the research design, including without knowledge of question, interview situation, subjects selection and interviewer behavior, the quality of the results is difficult to assess. Specifically, the opinion research also encounters the following problems:
- Just when you want to draw conclusions from an opinion on conduct or action, often shows that the expressed opinion of the respondents deviated on a behavior or action from the actual.
- When perceived as socially undesirable opinions / behaviors a part of the respondents tend to give these not admit or advanced responses (psychological rationalization ). When standardized written surveys the " yes- saying tendency " plays a distorting role. In addition, only the questions are answered - " associations, unconscious or preconscious of subjects remains undiscovered ."
- Both by the behavior of the interviewer ( interviewer bias) as well as by suggestive questions it is possible to provoke certain responses. This is to be avoided though, but is in practice regularly.
- For clients who wish to obtain tend to confirm a preconceived opinion, the kind of question as well as the selection of subjects are used to favor certain results.
- With direct, personal interviews, it can happen that often working in precarious jobs, interviewer "invent" the survey results to supplement their salary. This is possible if they are paid per questionnaire.
- In particular, in telephone interviews doubts about the representativeness are appropriate, as for the telephone interviews with badly planned surveys often only the so-called " head of the family " was questioned ( on the other hand is a correct approach, the birthday method). In addition, all those who fall from the sample out that such telephone surveys negative face and therefore do not even go into it. In particular, due to the increased incidence of telemarketing decreases probably the willingness to telephone surveys sections of the population. In addition deteriorates the fact that especially young people on a fixed line refrain ( and as a substitute use their mobile phone ), you may get the chance, representative statements of this population.
- The interviewed persons are often chosen to be very small in order to keep the evaluation work as low as possible. However, this in turn suffers the representativeness of the results. In particular, when sub-groups are formed, the actual number of respondents to persons may, at only 20, or even lower, which allows no more valid conclusions.
Example
In particular, in the general election in Germany in autumn 2005, it has been shown that the pollsters make clear incorrect predictions with errors of 5 % or more. In the specific case is still unclear where exactly was the cause, but it is assumed that the traditionally very high party loyalty in the Federal Republic of Germany gradually decreases to values as observed in the U.S. for some time, and that, therefore, the correction factors with which the primary data are weighted, gradually lose their validity and need to be adapted to the changing social perception.
Opinion research institutes
The pollsters try with different methodological approaches to compensate for these known sources of error through statistical interpretation.
The largest and best-known opinion research institutes in Germany are:
The largest and best-known opinion research institutes in Austria are:
- IFES - Institute for Empirical Social Research
- Integral market and opinion research
- Peter Hajek Public Opinion Strategies GmbH
- SORA
Known opinion research institutes in Switzerland are:
- DemoSCOPE Group
- Gfs.berne
Known opinion research institutes in France are:
- Institut français d' opinion publique, Paris
- TNS Sofres, Paris
Known opinion research institutes in the UK are:
- TNS BMRB
- YouGov, London
Applications
Opinion research is used in politics, the media, the economy and social research often.
In politics, opinion research is used on the one hand in the choice of research, on the other hand in the survey on specific policy issues. Democracy theory, this often encounters criticism: The orientation of the Policy at the polling cave traditional and proven forms of political decision-making from. The orientation of the policy on the collected by uncertain concepts fickle opinion would prefer populist instead of content more correct solutions.
The economy uses the opinion research, especially in the field of market research.