Yesterday, I was conducting an experiment that took a while to complete. That's why I skipped a day. I was looking into Google searches for Obama and McCain over the last three days available (10-16 through 10-18) on Google Insights for Search, and well split up by US state plus the District of Columbia. I experimented with different analyses and this is what I came up with:
The states are displayed alphabetically. Let's give them ranked according to Obama share too:
The interest in Obama was the highest in Missouri, then came Nevada, Wyoming, and so on. The lowest Google-popularity for Obama or highest one for McCain in this three-day time span was in Delaware, then North Dakota and Massachusetts. So did this mean that Obama was ahead in Missouri but behind in Delaware, opinion-poll-wise? Of course not. I think these graphs show a snapshot of the attention being paid to the two main presidential candidates, e.g., on October 18, Obama held a 100,000-people rally in St. Louis, MO. What is measured is more something like the "buzz," how much the candidates are being talked about. You might say that it reflected the cumulative effect of campaigning: TV, radio, robocalls, appearances, mail, etc.
The analysis also yielded the top cities where people were looking for information on the candidates during the sam time span:
This confirmed my theory that the huge St. Louis Obama rally spiked Missouri: St. Louis was no. 1 for Obama. Second came Houston and then Tampa. McCain had the highest score in Boston, then San Francisco and Seattle. It seemed to me that these had more to do with Obama being so far ahead that people weren't thinking about the candidates much anymore.
12 hours ago