SPON survey: Union in Europe better than in the Bund

The Union has launched its European election campaign this weekend. Now the new CDU boss Annegret Kramp Karrenbauer must deliver results. At the moment, however, the figures still look good. This is shown by representative surveys by the online polling agency Civey for SPIEGEL ONLINE.

If this European election Sunday would be CDU and CSU to just under 32 percent. Four weeks before Election Day, the Union is still a long way from its result of five years ago. At that time, the Union landed at 35.3 percent.

After all, the Union is on the European question in the tendency is better than in the classic Sunday question to the general election. The SPD is different: It scores just as badly on the European question and stands at 16 percent - just behind the Greens .

The other parties also weaken in comparison to the Sunday question also: The AfD comes to the poll, according to just over 11 percent, the left is just under 7 percent, the FDP at about 8 percent.

In the Sunday question to the federal election CDU and CSU remain under 30 percent. The Greens and the SPD are also in second place here, with the ecopartei's advantage over the Social Democrats tending to be greater than on the European question. The Greens come to just over 18 percent.

The AfD would currently come to around 13 percent in the federal government. Left and FDP are close to each other at about 9 percent, with the FDP has tended to increase since the weekend. Perhaps the FDP party congress has mobilized a few liberal voters.

How does the Civey method work?

The opinion research institute Civey works with a multi-level fully automated procedure. All representative real-time surveys are played in a Germany-wide network of more than 20,000 websites ("Riversampling"), so it is not only users of SPIEGEL ONLINE interviewed. Anyone can participate in the surveys online and will be included in the representative result with their answers if they have registered. From these users, Civey draws a quoted sample that ensures that it matches the population, for example, in terms of age, gender and population density. Finally, in a third step, the results are weighted by other attendees' socio-demographic factors and attitudes to correct distortions and prevent manipulation. More information can be found in the Civey FAQ.

Why is a registration necessary?

The registration helps to weigh the answers, thus allowing a result for the surveys, which is representative of the voting population in Germany. Each participant is asked for their gender, year of birth and place of residence. After that everyone can give their opinion in further surveys on different topics.

How do the results become representative?

The answer of each participant is weighted so that the result of a survey is representative of the population. For the Sunday question and the government monitor, this population comprises the population entitled to vote in Germany. The weighting is done fully automatically on the basis of the personal details at the registration as well as the history of earlier answers of a user. More methodological details can be found in the Civey whitepaper.

Will you reach enough participants online?

Opinion polls are usually conducted by phone or online. The significance of the results depends on how many people can be reached and how many actually participate in a survey when they are approached. Internet connections and landline connections are currently about equally widespread in Germany - with about 90 percent of households, mobile phones even 95 percent. The willingness to participate in all methods in the single-digit percentage range, especially experts estimate it for telephone surveys.
Thus, in both methods there is a group of people that can not be reached because they either have no connection to the respective network or do not want to participate in the survey. Therefore, a significant number of people must always be approached for a meaningful result. Civey surveys are currently in addition to SPIEGEL ONLINE in more than 20,000 other websites involved, including various media. This ensures that as many populations as possible can be reached.

How do I recognize the quality of a result?

Until the result of a survey becomes representative, enough different people have to participate. Whether this is already successful, makes Civey transparent, in that for each survey result a statistical error probability is specified. The number of participants and the interview time are also published for each survey.

What does it mean when the colored areas in the graphics overlap?

In our graphs, the statistical error is shown as a colored interval. This interval shows the uncertainty associated with a poll score. For example, on the Sunday question, one can not say exactly how many percent a party would get in a poll, but specify an interval where the outcome is likely to be. If the intervals of two survey values ​​overlap, then strictly speaking no statements about the difference can be made. For the Sunday question this means: If the poll numbers of two parties are so close together that overlap their error intervals, it can not be derived from which would currently perform better in the election.

What happens with my data?

The personal data of the users are stored encrypted on German servers and remain secret. Civey employees use only user IDs for reporting and can not associate users with their votes. The main purpose of the users' personal information is to weigh the answers and to ensure that the surveys are not manipulated. To prevent this, Civey uses both statistical and technical methods. In addition, Civey works with external partners who create audiences. Only when users have accepted the privacy policy of both Civey and an external partner, may your responses be used by the Partner to model those audiences. However, a partner does not receive information about your political and religious attitudes as well as those with which you can be identified. Civey users are also not ads based on their answers. You may object to the distribution to partners at any time here as a logged in user. More information about privacy at Civey can be found here.

Who is behind Civey surveys?

At this point, readers in the app and on the mobile / stationary website have the opportunity to participate in a representative Civey survey. Civey is an online opinion research institute based in Berlin. To compile its representative surveys, the software of the company, founded in 2015, merges websites into a nationwide survey network. In addition to SPIEGEL ONLINE include, among other things, the "Tagesspiegel", "World", "Wirtschaftswoche" and "Rheinische Post". Civey was funded by the ProFit funding program of Investitionsbank Berlin and the European Regional Development Fund.