Sri Lanka went through a rough patch in politics following the removal of the prime minister on October 26, 2018. Following this incident, certain MPs jumped from one party to another making the entire philosophy of politics a joke. Having seen this, I decided to draw similarities of the Sri Lankan parliament with Principal Component Analysis (PCA), and the politics to quantitative research.
Politics is like quantitative research. Thousands of people provide data, just like votes. One or two folks will decide what they need to do with those data. Ultimately, you are just a piece of data. Once you have given the data, the researchers can manipulate them in any way they like. They only see the bigger picture. End of the day, both these people receive some bogus reward. So, are we deceived? Well, yes. Media channels are already in the debriefing process.
***********************************************************************************************************************************
Parliament is like principal component analysis (PCA). Initial dimension reduction does not demonstrate the expected number of loadings for some principal components. Let's assume the components are the parties and the loadings are the ministers that help you to determine which party has the majority. So each time you do a rotation, some loadings (ministers) like Wasantha Senanayaka and Vadivel Suresh keep on jumping from one component (party) to another until the required majority (expected factor structure) is determined. In unusual circumstances where some small components (like JVP) are not so significant, due to some shared similarities they are considered with the component with the highest variance or similarity (UNP). Haha. However, the best method of rotation would further extract loadings (Manusha, Fowzie, Piyasena Gamage) towards the component with the highest eigenvalue (majority). Now you can keep it like this, or you could continue to rotate further.
Dumb joke, but funny if you into the whole factor analysis thing.