Verfijn uw zoekresultaat

Zoekresultaat: 2 artikelen

x
Artikel

Access_open In de schaduw, uit de schaduw

Oorsprong, aard en mogelijkheden van schaduwverkiezingen of exit polls

Tijdschrift Bestuurs­wetenschappen, Aflevering 4 2018
Auteurs Prof. dr. Jelke Bethlehem en Prof. dr. Joop van Holsteyn
SamenvattingAuteursinformatie

    There is a lot of polling in the Netherlands, especially in the run-up to elections. The assessment of future voting behavior in the run-up to elections is inherently difficult, because many voters do not know in advance whether they will vote, let alone for which party. There is therefore constant debate about the quality of these surveys. However, there are also polls that are not held prior to elections, but on election day instead. They are called exit polls or shadow elections. The sample consists of voters who actually visited the polling station and cast their vote. In this article the authors emphasize the nature and useful and interesting role of exit polls. Exit polls are an important tool for making an accurate prognosis of the results shortly after the closing of the ballot boxes. Secondly, an exit poll can provide further insight into electoral gains and losses, and thus counteract unfounded speculation. After all, the data collected form an empirical source for a first analysis of the outcome and electoral behavior. All in all, the exit poll is a relatively easy-to-organize and attractive ingredient for a results evening. Confusing pre-election polls with exit polls probably does not do justice to the higher quality of exit polls in terms of prognosis. The article explains where exit polls differ from pre-election polls and what the most important choices are when setting up such a poll; it also shows that a well-designed exit poll is accurate and has adds value to a results evening. The authors give practical examples in their argument and discuss the exit poll that was organized in Leiden at the council elections of 21 March 2018.


Prof. dr. Jelke Bethlehem
Prof. dr. J.G. Bethlehem is bijzonder hoogleraar in de survey-methodologie aan het Instituut voor Politieke Wetenschap van de Universiteit Leiden. Hij was tevens senior methodologisch adviseur bij het Centraal Bureau voor de Statistiek in Den Haag.

Prof. dr. Joop van Holsteyn
Prof. dr. J.J.M. van Holsteyn is hoogleraar politiek gedrag en onderzoeksmethoden aan het Instituut voor Politieke Wetenschap van de Universiteit Leiden.

    Polling is being done a great deal in the Netherlands, especially during election campaigns when market researchers sometimes present new polls every day. The national government also takes polls that are often larger and more complicated than the quick and small polls conducted by market research agencies. They are often called surveys, and they gather information on the state of affairs in society. That information can become the basis for new policies. Local governments also take polls, although on a smaller scale than national government. Dutch municipalities have a tradition of organizing omnibus surveys in which (as the name indicates) several subjects can be addressed. Nowadays many ‘omnibus surveys have been replaced by ‘citizen panels’. One thing all these polls and surveys have in common is that they are based on random samples of the population and statements are made about the population as a whole based on these samples. Such generalizations are only possible if the sample is drawn using by random sampling methods. This article describes good and bad polling. This is illustrated using a unique example: the research into the opinion of the inhabitants of Alphen aan den Rijn, a Dutch municipality, on Sunday shopping. At the same time, and using the same questionnaire, three different polls were carried out. This example makes clear that the wrong sample can lead to incorrect conclusions and maybe to incorrect policy decisions.


Prof. dr. Jelke Bethlehem
Prof. dr. J.G. Bethlehem is bijzonder hoogleraar in de survey-methodologie aan het Instituut voor Politieke Wetenschap van de Universiteit Leiden. Hij is tevens senior methodologisch adviseur bij het Centraal Bureau voor de Statistiek in Den Haag.
Interface Showing Amount
U kunt door de volledige tekst zoeken naar alle artikelen door uw zoekterm in het zoekveld in te vullen. Als u op de knop 'Zoek' heeft geklikt komt u op de zoekresultatenpagina met filters, die u helpen om snel bij het door u gezochte artikel te komen. Er zijn op dit moment twee filters: rubriek en jaar.