First, let's look at the picture. (To see them in full size, just click on them) So what you see here is a REG recording realised with the OnlineFieldREG service of Psyleron. The beginning is at 10h30 EST (local time, UTC -5. I'll give all times in a 24h format and with this timezone) and the end 4 hours later. So overall, it's not really a special recording, it's quite close to the chance level at the end but it's inside the "normal" zone. But that's something logical, because the event we want to look for took place only a part of this time. Precisely, it takes place between 12h05 and 12h26:05. There is a picture of the full session with annotations just under this paragraph so you can see more precisely when the event is situated on the graph. And it's here that it becomes interesting :)
As you can see, the inauguration itself, the vow and the adress, is an especially "positive" moment. I reported this part of the graph at the origin to compare it to the "chance level", and you can see that it's clearly positive. Let's talk more precisely of this section of the recording.
For this analysis, I used the official video of the event, released on the website of the White House. At the time of my analysis it was on another temporary website (http://www.pic2009.org), which one don't exist anymore. On the blog, I got the official time of the vow, which is the beginning of the video : 12h05. So, it's great, because we have something official, but it's also bad because my REG recording is in seconds, not minutes so I don't know if it is really 12h05:00 or 12h05:59... So as I'm left with a 60s uncertainty, I put error squares on the graph.
There is a lot of things in this graph so I'm gonna explain it all. The axes are the same as those of the full version. The original curve is the one with low opacity. I shift it to 0 to see the cumulative deviation only for this part of the graph. There is 5 points on the graph that are shown and explained at the bottom of the graph. These are the 5 key moments of the vow and the adress. The green lines that marks these moments comes with a green rectangle on their left, which is the uncertainty I spoke before. The pink lines are moments during the adress where there was a lot of applause. I thought it could be interesting to see if we can have a correlation with that also.
So, what can we say about this graph? I think it's not bad. It's not perfectly in sync with the facts but the trend is good. Plus the fact that it is over the "chance level" nearly anytime. But the bad thing is that we have those uncertainties about the exact time of the event, which didn't really help our analysis second by second. But that's not all. We have another uncertainty : did the REG device and the White house have the exact same time or less than 1s of difference between them? We can't really know so maybe the graph as to be shifted a little to be right. Let's try it !
So for this version of the analysis, I took both the fact that we have uncertainties about the precise time of the beginning of the recording and about the time difference between the White House and the REG. So what I did was to take all the lines, green and pink, and shift them left or right to see if I could find a place where most of the lines correspond to special points in the graph. The result are visible on the picture, so yes, we can have something looking better simply by shifting the graph of 83s, 1min23s. So, if we postulate that there is this difference of timing between the REG and the real events (which I recall you that we have 60s of uncertainty about them), then we have something really intersting :
Here, we can really determine the adress, with a special point at the beginning with a big rise, and the highest point of this portion of the graph with the end of the adress. So sure, that's prettier, but could it be real and explainable? I think that yes. From a lot of studies I saw about REG, they seems to measure "mass thought coherence". Normally, the graph would look flat over a certain period of time. If there is a rise in coherence, then the curve will rise, what I call "being positive". If, on the contrary, there is less coherence, then the curve will go down, "negative". Applied to our particular case, I think that we can postulate that moments where people will listen carefully to Barak Obama, then the curve will rise. When people will start to think to something else, like when there is applause or the things he says make them think, the curve will go down. In normal times, there is not enough focus to see "second" details like that, but I think that when we have a worldwide event like that, with billions of thought directed toward the same thing, the effect can be that precise.
If we look at the graph, we can see clearly that when the adress began, there is a very sharp rise in coherence. And of course that is what happened at that time, people began to listen to him. At every "applause time", the rise slow, even reverse. And at the end, when everybody knows that the adress is finished, there is a sharp fall, near a mirrored curve of the beginning. So, I can't really know if this is the exact timing, but it is a likely possibility, and by the analysis we can do with it, a very likely one.
Now let's get back to the full picture, because something else is interesting. As you can see on the official schedule, the inauguration really begun at 11h30. So I took the part from 11h30 to 12h30, and put it at the origin of the graph to show how it compare to chance levels. The graph is very "positive". I don't know if my calculations are correct but I found values of Z=2.66 and p<0.0039. That seems pretty good. And you can see that from 12h30 to 13h00, it's quite stable. Between 13h00 and 13h30, it begins to rise again, but a little less than before. After that, there is a low point at 14h11, then it rises until the end.