deletions | additions       diff --git a/Discussion - Sources of Bias.tex b/Discussion - Sources of Bias.tex  index 1ec0fa2..e69d816 100644  --- a/Discussion - Sources of Bias.tex  +++ b/Discussion - Sources of Bias.tex 
...
When compared to major polling companies and their ongoing polls published in the media, it seems that several parties are under-represented while others are over-represented, even after our standardization process.  Some notable examples are that according to our prediction:  \begin{itemize}  \item  The left-wing party Meretz gets 9 seats (4-5 seats in other major polls). \end{itemize}  \begin{itemize}  \item  The extreme-right-wing party Yahad gets 9 seats (4-5 in other major polls). \end{itemize}  \begin{itemize}  \item  The separadic-orthodox party Shas gets 0 seats (8 in other major polls) \end{itemize}  \begin{itemize}  \item  The right-wing party Israel-Beytenu gets 0 seats (5 in other major polls) \end{itemize}  One possible source of bias is the influence of abstention (non-voting). Our model does not incorporate a mechanism to assess changes in turn-out. While we did ask our respondents whether they abstained in 2013, it would be naive to assume that they represent the non-voting population. A non voter is presumably indifferent and would not participate in our poll. Those who do participate, probably intend to vote in 2015. We could therefore easily reach the false conclusion that turn-out will increase to nearly 100\% giving those users who reported abstention in 2013 unreasonably high weights. Eventually we chose to ignore possible changes in abstention, implicitly assuming that the voting population is constant and that we need only to infer if and how they will change their vote. In particular, there are media reports that turn-out will increase dramatically in the Arab population, which could increase the number of seats for the Arab Union.  Other possible sources of bias are several demographic variables that we were not controlling for (see section 2.2). Since we are already controlling for the 2013 election results (i.e. our weighted sample 'agrees' with the actual 2013 results), the following question is of relevance to the sources of bias:  In what way are our respondents \textbf{that voted for party i in 2013} different then the actual voting population \textbf{that voted for party i in 2013}. Some variables that may have played a role in biasing our sample could be:  1. \begin{enumerate}  \item  Sex. It is possible that females 'changed their minds' in a manner different then men. Ha'Mahane Ha'Zioni has a lot of women in their list, including Livni who is set to become prime minister through rotation with Herzog. As an example, females who voted for Yesh Atid in 2013 just might be more likely to switch to Ha'Mahane Ha'Zioni than men who voted for Yesh Atid in 2013. 2. \end{enumerate}  \begin{enumerate}  \item  Age. It is not unreasonable to assume that young voters are more likely to vote for new, small, niche or extreme parties than are older voters. For example, older voters who voted for Likud in 2013 are more likely to stick with Likud than are youngers voters, who are in turn more likely to switch to Yahad or Kulanu. This same bias could explain the high number of seats projected for Meretz and the low number of seats projected for Israel Beytenu. \end{enumerate}  Lastly, yet another possible cause for bias in the app data is that 2013 votes were only recorded after February 13th, and only in Android device. Therefore, the 2013 dataset contains only ~2400 votes, roughly a third of our entire sample.