Election 2016 Intro: Counter Arguments and Potential Weaknesses
As a general disclaimer, some weaknesses with the general analysis should be highlighted:
1. Media Bias and Social Desirability Bias – these go hand-in-hand when it comes to national political campaigns so if you disregard an inherent media bias that is weighted against Trump, you will likely have a tough time buying the Social Desirability Bias argument. More research is needed in order to better quantify media bias and its impact on elections. Multiple assumptions are made along the way that could prove incorrect not least of which is that the media is generally anti-Trump.
2. New Data and Analysis Methods – using social media and on-line activity analysis is relatively new and not that well tested in that its track record, though very good, is very short. Without a longer track record showing potential pitfalls, there are bound to be issues that could come up to throw off some of the conclusions.
3. Lack of Access to Poll Data – in order to better analyze how polls could be erroring it would be best to have access to complete poll data and methodology, instead this analysis is based on publically available data and information. More specifically, it would be much better to have a transparent understanding of the re-weighting process which is believed to be producing inherent pro-Clinton skews in the polls.
4. Focus on Perception – much of the analysis focuses on how the electorate perceives reality and not on reality itself as perception is much more important for intangible things such as feeling enough indirect social pressure to make an individual change a poll answer. Perceptions are easy to discuss but difficult to prove.
5. Making too Many Financial Analogies – as previously stated, my background is heavy on financial / investment / social media analysis and much of the ideas that I bring to this political analysis are rooted in the idea that analytical techniques that have worked when analyzing investments will carry over to politics.
6. Not a Political Analyst – there have got to be some nuances that have been overlooked due to inexperience with political analysis.
7. Inappropriate Historical Analysis – throughout the analysis, historical election examples and data are used to bolster some observations. For instance, the 1980 US Presidential election is referred to on various occasions as a potential similar case worthy of analysis, but such comparison could easily be proven wrong and shown to stretch similarities between the elections.
With this as a backdrop, please enjoy the posts. For comments or specific questions regarding the social media and on-line activity analysis, please use kevin@zettacap.com. For all other inquiries, feel free to contact me via kjcoogan@yahoo.com or on twitter at @kjcoogan.