[ad_1]
The survey was conducted on behalf of SRG SSR by the research institute gfs.bern between September 2 and 10, 2020. In total, responses from 17,909 eligible voters were taken into account for the evaluation.
1,407 people with voting rights residing in Switzerland were interviewed by telephone: 701 people from German-speaking Switzerland, 405 from French Switzerland and 301 from Italian Switzerland. The interviews were conducted by landline and cell phone.
This sample is weighted according to linguistic region and is representative of Swiss voters. The statistical error is ± 2.7 percentage points. With 1407 respondents and a result of 50 percent, the effective value is between 47.3 and 52.7 percent with a probability of 95 percent. Smaller deviations are more likely, larger ones are less likely.
Online survey
In addition, several thousand people were interviewed online. Once the data has been cleaned and verified, information from 16,502 eligible voters could be used for analysis.
The distribution of online respondents by language region is as follows: 14,065 in German-speaking Switzerland, 2,864 in French-speaking Switzerland, and 980 in Italian-speaking Switzerland. The online survey was conducted through the SRG Medien web portals as a so-called opt-in (participatory survey).
This collection of online samples was not random and the resulting sample is not representative. For example, the online survey included fewer retirees than youth and more men than women.
Therefore, the gfs.bern institute weighed the responses: Sample biases were counteracted by statistical weighting procedures and representativeness was optimized.
How do you ask?
The surveyed voters each had five possible responses: “definitely in favor”, “rather in favor”, “don’t know / no answer”, “definitely against” and “rather against”.
For a simplified representation in the article, the answers “definitely in favor” and “rather in favor” were added in most cases; the “definitely against” and “instead of against” responses were treated accordingly.
Specifically, the following question was posed: “Regardless of the certainty you have of participating in this referendum: if the bill were voted tomorrow, would you be in favor, in favor, against or definitely against?
Surveys are instant
The gfs.bern research institute publishes two polls for the September 27, 2020 vote. The study authors emphasize that the results are not an early voting result, but a snapshot at the time of the poll. However, trend statements are possible in at least two surveys.
Detailed information on the type of survey and the interpretation of the results can be found on the institute website gfs.bern, Link opens in a new window.
