As many of you students know Monday one of the last debates was held in the Erasmus Aula at our university between the currently biggest expected parties VVD and PVV. It was a wonderful debate where Mark Rutte (VVD), according to the media, slowly but surely got the upper hand over Geert Wilders (PVV). This debate was a precursor to the elections of today, the day that will decide who will rule our country for four years.
Predictions say VVD will be the biggest party again, but by experience from other elections for example in the United States, we know that predictions can we misleading. The statistics interpreters from I Hate Statistics, who want to help everyone to understand data and statistics, made the interactive explainer 'Peil je Wijzer' (Dutch only) below where you can become a surveyor yourself and see that predictions and exit polls are not accurate.
Click here for the Peil je Wijzer application
We will compare two polls made by different institutes with each other and show you how they can be so different. The first election poll is provided by IPSOS NEDERLAND and the second is by I&O Research.
Even though it is a small difference, it can have a huge impact on the voting behaviour
As you can see there is a minor difference. Even though it is a small difference, it can have a huge impact on the voting behaviour, because some people are ‘’tactically voting’’ as a result of the polls. But we are econometricians so we decided to make our own prediction using our knowledge. We decided that the A/F-test would be the most suitable test for our prediction. A/F-test splits a group randomly in two and gives those two groups the same test/sample but implementing the FAECTOR Factor: How much green does a specific party have. When implementing this A/F-test with the FAECTOR Factor, the results are as follows:
As seen above, we predict a victory from the GroenLinks, D66 and CDA party, this maybe has to do with the beautiful green color that the parties and our well loved association FAECTOR share. Another reason could be that the chairman of GroenLinks, Jesse Klaver, seems to be a popular politician under the feminine voters.
Tonight we will know for sure who won the elections and if our predictions were right. We shall see if econometricians are not only good in predicting economical and financial subjects, but also in political subjects.