This is an interesting study of anomalies regarding exit polls regarding several hot button election issues and politicians. The work looks like it was done using good methodology including voluntary blind exit polling.
The work covers several issues as a means of showing consistency among polling results and the work shows variations from the expected results outside what would be acceptable differences. Due to the fact that proprietary knowledge now trumps transparency in vote counting the work can't observe actual vote results but is dependent on exit polls.
One of the areas covered in the many areas in the country that this study includes was the New Hampshire vote.
What strikes me is the fact that the disparities are focused to favor the conservative agenda. That may be a factor in what was chosen to be polled but the author did include a variety of issues as to produce a more accurate observation. The Election Defense Alliance has a fairly good summary of the work as well as a link to the original pdf.
If you are interested in actively countering the results of this proposition in regards to the questionable vote count there is allready a site set up to handle the effort needed to bring the voter back into the political process.
I hope the mainstream media covers this but in the current political climate that seems unlikely. We as voters have to show we do want our votes counted and results like these are the reason why we need transparency in our vote counting methodology.